Multiple image segmentation and / or multiple dynamic spectral acquisition for materials and mineral classification

Abstract

To provide a system configured for material analysis and mineralogy.SOLUTION: A scanning microscope system comprising a first detector and a second detector, and a data processing system comprising a data storage component and a first spectral analysis component, where the data storage component is configured to provide an image of a sample or section thereof. A second detector is configured to detect a second emission from a second scan location, the data storage component is configured to provide a first spectrum, and the first spectrum is based on the second emission, the first spectral analysis component calculates a confidence score and is configured to select a second scan location associated with the first spectrum below a threshold. The second detector is configured to detect the second emission and the data storage component is configured to provide a new image defining the second scan location and to determine a new second scan location in a new image.SELECTED DRAWING: Figure 8

Description

[Technical Field]

[0001] The present invention relates to the field of spectroscopy and image analysis. The present invention further relates to determining the properties of a sample or sections thereof, for example by multiple image segmentation and / or multiple dynamic spectral acquisitions. [Background technology]

[0002] Materials research, which involves characterizing probe properties (e.g., structure, topography, chemical composition) in the microscopic and nanoscopic regimes, can be performed through the implementation of scanning microscopy systems, such as scanning electron microscopes (SEMs). SEMs are configured to scan the surface of a sample with a primary beam (i.e., an electron beam) and acquire images of the sample based on various types of emissions, e.g., backscattered, transmitted, or secondary electron emissions. These emissions result from the interaction of the electron beam with particles (e.g., atoms) of the sample. In the case of mineralogy research, the sample consists of thousands of mineral grains within particles embedded in an epoxy matrix.

[0003] Backscattered electrons (BSE) originate from a primary electron beam, which, as the name suggests, is reflected (i.e., exits) from the sample via elastic scattering off sample atoms. The number of backscattered electrons at each scan location on the sample depends on the atomic number of the chemical element (e.g., mineral element) located at the corresponding scan location. Therefore, intensity variations (e.g., gray level variations) in the BSE image indicate compositional variations within the sample.

[0004] Along with the emission of backscattered electrons, X-ray emission can also result from the interaction of the primary beam with the sample. In particular, characteristic X-rays are emitted when the primary electrons eject electrons from the inner shells of the sample atoms, creating electron holes. This electron hole is then filled by another electron from the outer shell of the atom through the emission of an X-ray photon. The energy of the X-ray photon corresponds to the energy difference between the outer and inner shells. Thus, the emitted X-rays have energies specific to the corresponding chemical element, and their detection reveals the chemical composition of the sample. To detect X-ray emission, SEMs are equipped with an X-ray spectrometer configured to measure the number of detected X-rays with respect to their energy (energy dispersive spectrometer, EDS) or their wavelength (wavelength dispersive spectrometer, WDS).

[0005] Materials analysis (e.g., mineralogy classification) typically involves combining backscattered electron imaging processing with the application of X-ray spectroscopy. However, while X-ray acquisition takes a few milliseconds per scan location, acquiring BSE at each scan location can be 3-4 orders of magnitude faster. Therefore, obtaining compositional information for an entire sample based on X-ray detection from tens, hundreds, or even thousands of scan locations can be very time-consuming, lasting from minutes to hours.

[0006] For this reason, a common measurement mode for materials research (e.g., mineralogy classification) disclosed in EP 2546638 B1 is to reduce the number of scan locations for X-ray detection. This is done by acquiring a high-resolution BSE image and segmenting the image to identify areas (e.g., mineral grains) with the same intensity and therefore the same chemical composition (e.g., mineral composition). Only one scan location is determined for each identified mineral grain. Next, the primary beam is positioned at the scan location of each identified grain to detect the corresponding X-ray emission and obtain a respective X-ray spectrum. Thus, mineral grains are first distinguished based on their different intensities (e.g., gray level intensities) on the BSE image, and then their chemical compositions are classified in their respective X-ray spectra.

[0007] To achieve fast acquisition and maximize acquisition throughput, the current solution is to perform BSE image segmentation in parallel with the X-ray acquisition, as disclosed in another recently filed application by FEI, EP20217928A1, which is incorporated herein by reference.

[0008] While most parts within a sample (e.g., mineral grains) are reliably identifiable (distinguishable) based on X-ray spectra acquired within a few milliseconds (5–10 ms), there is a small subset of minerals that produce X-ray spectra similar to other minerals. For example, various smectites are easily confused with illite. Despite the fact that illite has a distinctive potassium spectral line at 3.2 keV, structural rearrangements within the crystal structure (exchange of K and HO) can change the height of the potassium line. In this way, the illite spectrum loses its "uniqueness" and becomes indistinguishable from the smectite spectrum. Other examples of mineral grains that are not easily identifiable include iron oxide minerals, hematite and magnetite (Fe2O3 and Fe3O4), and various copper sulfides (CuS, Cu2S, etc.). Correctly identifying these specific minerals can be of great economic importance in industry. For example, mining companies are paid based on the specific grades of iron and / or copper contained in their extracted minerals. Therefore, knowing what iron oxides or copper sulfides are present in a deposit can have a considerable impact on a company's profits.

[0009] As a result, reliable differentiation of chemically similar moieties (e.g., mineral grains) requires the enhancement of the spectral quality (e.g., spectral resolution) of the corresponding X-ray spectra. An example of a current solution is to apply automated X-ray acquisition to all moieties (e.g., minerals) within a sample and increase the integration time (i.e., dwell time) by approximately 10 times or more (approximately 100 milliseconds). This allows for high photon counts (up to 20,000 photons per spectrum) in each mineral grain, subsequently improving spectral quality and helping to resolve all minerals. However, this approach results in low system throughput, slow processing, and over-acquisition of easily identifiable minerals, which would otherwise require much shorter dwell times (approximately 5 milliseconds). Therefore, this current approach is not commercially viable, as companies typically need to perform multiple measurements to extract all the necessary information from a sample of interest.

[0010] Material analysis (e.g., mineral classification) becomes particularly challenging when different parts (e.g., mineral grains) of similar composition are segmented into a single part (i.e., a single mineral phase) because they are associated with similar intensities on BSE images. The current solution to this problem is to manually locate parts (e.g., minerals) that are not easily distinguishable in high-contrast and brightness modes and then increase the contrast and brightness of the selected parts until two or more parts are visible. Re-acquiring images of the sample with these modified contrast and brightness values ​​and applying image segmentation and X-ray acquisition to the newly revealed parts (e.g., mineral grains) can help determine their corresponding chemical compositions (e.g., Fe2O4 and Fe3O4). However, with this workaround, some other mineral grains in the sample may be considered background because they are depicted as too bright or too dark by the modified contrast and brightness settings. Therefore, this approach can lead to "cut-out" parts (e.g., mineral grains) that can introduce additional errors into statistical reports on the properties of the sample of interest, such as average sample content. Summary of the Invention

[0011] The present invention seeks to overcome or at least mitigate the shortcomings and disadvantages of the prior art. More particularly, it is an object of the present invention to provide improved methods, systems, and computer program products for materials and mineral analysis.

[0012] An optional object of the present invention is to provide a system and method for determining properties (e.g., chemical composition) of a sample and / or sections thereof. In particular, an optional object of the present invention is to enable image segmentation and adjustable X-ray acquisition. Another optional object of the present invention is to enable secondary dynamic X-ray acquisition and / or secondary image segmentation for selected sections of the sample.

[0013] In a first embodiment, a system including a scanning microscope system and a data processing system is disclosed. The system may be configured to provide at least one or more images of a sample and / or a section thereof based on first emissions detected within a first dwell period from a plurality of first scan locations. The system may further be configured to detect second emissions during a second dwell period from at least one or more second scan locations in at least one region of the at least one image, each second scan location associated with a portion of the corresponding region. The system may further be configured to provide at least one or more first spectra, each first spectrum based on second emissions detected at each of the second scan locations in the at least one region. The system may also be configured to calculate a reliability score for all first spectra and select second scan locations associated with first spectra whose reliability scores are below a threshold. Further, the system may be configured to detect a second emission from at least one of the selected second scan locations during a third dwell period, and / or provide at least one or more new images demarcating portions associated with the selected second scan locations, and determine a new second scan location in the corresponding new image via the modified contrast and brightness values ​​for the at least one image.

[0014] The term "image" is intended to include a two-dimensional grid, which may include at least one or more portions. Each portion is characterized by its coordinates and a value (color and / or intensity). Thus, an image may also refer to a visual representation of a sample in terms of varying colors and / or varying intensities. For example, an image may include variations in intensity of the same color, such as variations in gray level. Furthermore, each portion in an image may correspond to a point on the sample (e.g., a scan point). An image portion may be, for example, a pixel or may include multiple pixels.

[0015] Furthermore, the term "mask" is intended to include, for example, binary images containing black and white portions. Portions of one color and / or intensity (e.g., white portions) may be used to mark sections of an image for further processing. However, the term mask may also refer to the marked sections (e.g., white portions) of an image.

[0016] The term "spectrum" is intended to include the distribution function of a physical quantity (e.g., energy or frequency). The measurement of a quantity can be, for example, the intensity, abundance, velocity, or flux of each quantity value. A spectrum can refer to a discrete spectrum, which can include a set of discrete spectral lines at different energy values. The peak of each spectral line at the center of the corresponding line can correspond to the maximum number of detected photons (i.e., peak intensity) across the respective linewidth. Detected photons can further refer to detected X-ray photons. Each spectral line can correspond to an electronic transition of a chemical element, and the energy value of each electronic transition can be unique for the corresponding chemical element. A spectrum can also refer to a continuous spectrum, which can refer to the intensity distribution over a range of continuous energy values. However, intensity can also be plotted with respect to the corresponding wavelength, frequency, or wavenumber.

[0017] The term "particle" is intended to include a particle in a sample. A particle may correspond to a region. The term "region" may refer to a region of a sample that corresponds to a particle or portion thereof, for example, if only a section of the sample is imaged and / or processed and that section includes only a portion of the particle. The term "region" may also refer to a portion of an image, where that portion corresponds to a particle in the sample.

[0018] The term "mineral grain" is intended to include a mineral grain within a particle located within the sample. A mineral grain may correspond to a portion. The term "portion" may refer to a portion of a particle that corresponds to a mineral grain or a portion thereof, for example, if only a section of the particle is imaged and / or processed and the section includes only a portion of the mineral grain. The term "portion" may also refer to a portion of an image, where the portion corresponds to a mineral grain within a particle located within the sample.

[0019] Whenever x, y, and / or z coordinates or directions are used within this disclosure, the z direction may be vertical, i.e., orthogonal to the ground. The x and y directions may be orthogonal to each other and to the z direction, i.e., they may be horizontal. The coordinates may form a Cartesian coordinate system.

[0020] The term "scan location" is intended to include the location of a scan point within a sample, where the location is given by (x, y) coordinates with respect to the internal coordinate system of the sample and / or image.

[0021] Additionally, the terms "second scanning location," "region," image, spectrum, and any other term in the plural are used with the plural form of the verb for reasons of clarity and conciseness. However, these statements are intended to encompass at least one second scanning location, at least one region, etc.

[0022] In this disclosure, the term "time interval" is intended to include a period defined between two fixed times / events. Those skilled in the art will readily understand that two time intervals defined by the limits (t1, t2) and (t1', t2') of lengths w and w', respectively, where t1 ≦ t1', overlap if the following condition is met: w + w' > t2' - t1. A first method step performed during the duration of a first time interval and a second method step performed within a second time interval are intended to include parallel steps if the first and second time intervals overlap. Thus, two method steps are considered parallel if there is a partial or complete overlap of the corresponding time intervals.

[0023] The term "dataset" is intended to comprise a collection of data. The term "dataset" may also refer to a list of (x, y) coordinates of corresponding second scan locations. A synonym for "dataset" herein is "group."

[0024] The scanning microscope system may include a first detector, which may be configured to detect a first emission from the first scanning location.

[0025] The first detector may include a backscattered electron detector.

[0026] The scanning microscope system may include a second detector, which may be configured to detect a second emission from the second scanning location.

[0027] The second detector may include an X-ray detector.

[0028] A scanning microscope system may be configured to focus a beam of charged particles (such as electrons) to a scan point on a sample.

[0029] The scanning microscope system may be further configured to scan the beam of charged particles across multiple scan locations in one or two dimensions.

[0030] The scan location may correspond to the first scan location.

[0031] The scan location may correspond to a second scan location.

[0032] The data processing system may be configured to assign a two-dimensional coordinate system to the sample.

[0033] The data processing system may also be configured to assign a two-dimensional coordinate system of the sample to the at least one image.

[0034] Thus, as the beam of charged particles moves across a first scan location on the sample, the location of each moiety in the image can be tracked.

[0035] Assigning the same coordinate system of the sample to the image can be achieved using reference points of known coordinates, which can be built into the sample or the movable stage.

[0036] The results of scanning a beam of charged particles over a scan location on a sample may include interactions between the beam and the sample.

[0037] The result of the interaction can include a first and / or a second emission.

[0038] The first emission may include an emission of particles (such as backscattered electrons).

[0039] The second emission may include the emission of photons (such as X-ray photons).

[0040] The data processing system may be configured to generate at least one image based on the first emissions detected at each first scan location.

[0041] At least one image may correspond to a backscattered electron image.

[0042] At least one image may include contrast and brightness values.

[0043] Additionally, at least one image may show intensity variations between regions (and / or portions thereof) having different properties (such as chemical composition).

[0044] The intensity variation may include a gray level variation. In particular, a gray level image may include 256 levels of gray, with gray level values ​​ranging from 0 to 255.

[0045] In fact, the gray level intensity of an image (or a portion thereof) may be linearly related to the atomic number (eg, mean atomic number) of the corresponding section, as given by the following equation:

[0046] Formula 1: I=C * S+B,

[0047] where I represents the intensity of the gray level, S is a factor related to the atomic number (e.g., mean atomic number) of the section of interest (e.g., mineral grain) of the sample, and the C and B coefficients represent the contrast and brightness values ​​of the image. Therefore, if the C and B coefficients of an image are known, a user can derive the atomic number of the section of interest of the sample based on the corresponding gray level intensities of the image (see also Hardig (2002), "Mineral identification using a scanning electron microscope," Department of Metallurgical Engineering, University of Utah, Salt Lake City, Utah).

[0048] Each region of the at least one image may correspond to a particle in the sample.

[0049] Furthermore, each particle within the sample may contain at least one or more mineral grains.

[0050] The scanning microscope system, and in particular the first detector, may be configured to detect the first emission for the duration of the first dwell period at each first scan location.

[0051] The scanning microscope system, and in particular the second detector, may be configured to detect the second emission at each second scan location for the duration of the second dwell period.

[0052] The second dwell period may be longer than the first dwell period, e.g., the second dwell period may correspond to 8 milliseconds and the first dwell period may correspond to 1 μs.

[0053] The system may be configured to detect a first emission from a first scan location and a second emission from a second scan location at different time intervals, the different time intervals corresponding to non-overlapping time intervals.

[0054] The data processing system may comprise a data storage component, which may be configured to provide at least one image of the sample (or a section thereof).

[0055] The data processing system may include a first segmentation component, which may be configured to determine a second scan location of a region of the at least one image.

[0056] The first segmentation component may be further configured to determine each second scan location for the duration of the segmentation period.

[0057] The segmentation period may depend on image characteristics such as the resolution and magnification of at least one image (and / or section thereof).

[0058] The segmentation period may also depend on the size of the mineral grains and / or particles. Typical grain sizes range from at least 1 μm to a maximum of 500 μm.

[0059] Sample statistics such as particle size, number of grains per particle, and size distribution across samples have low statistical spread across the same sample and / or replicant samples, which can result in less variation in segmentation duration per second scan location.

[0060] The segmentation period may be equal to or less than the second dwell period, for example, the segmentation period may correspond to an average of 3 milliseconds and the second dwell period may correspond to 8 milliseconds.

[0061] However, the segmentation period may also be longer than the second dwell period, which may be the case if the image (and / or sections thereof) are highly resolved and / or enlarged and / or if the grain size is significantly large.

[0062] Generally, the resolution and magnification of at least one image may define a feature (e.g., pixel) size ranging from at least 10 nm to at most 1000 nm. For example, a smaller feature size, such as 10 nm, may indicate a higher resolution and / or a higher magnification. The smaller the feature size, the greater the number of features that represent the same part of the image (related to a mineral grain).

[0063] Indeed, the segmentation period can slow as the number of parts increases. For example, if a medium-sized (e.g., 50 μm) mineral grain is imaged at high resolution and / or high magnification (e.g., 20 nm part size), the resulting segmentation period can be very long (e.g., 10 times longer than the second dwell period).

[0064] The data processing system may include a pre-processing component.

[0065] The pre-processing component may be configured to apply a thresholding algorithm.

[0066] In particular, the thresholding algorithm may be configured to separate at least one image into background and foreground portions based on a threshold strength.

[0067] The background portion may comprise a background portion (eg, a pixel) that includes an intensity lower than a threshold intensity (eg, a dark gray and / or black portion).

[0068] The foreground portion may comprise a foreground portion (eg, a pixel) that includes an intensity equal to or greater than a threshold intensity (eg, a light gray and / or white portion).

[0069] Segmenting and / or removing background portions may include assigning the same color and / or intensity values ​​to the background portions (eg, black portions).

[0070] The foreground may include at least a portion of the image.

[0071] Additionally, the pre-processing component may be configured to determine boundaries of corresponding regions of the at least one image using a contour algorithm.

[0072] The contour algorithm may be configured to join adjacent portions along the boundaries of the corresponding regions into curves.

[0073] The adjacent portions along the border of the corresponding region may be surrounded by a background portion (eg, a black portion).

[0074] The preprocessing component may be further configured to apply a bounding box algorithm.

[0075] The bounding box algorithm may be configured to divide the at least one image into at least one or more sub-images based on the results of the contour algorithm.

[0076] A sub-image of at least one image may delimit a region.

[0077] The data processing system, and in particular the data storage component, may be configured to provide the sub-images.

[0078] The first segmentation component can be configured to correct for sub-image generation errors.

[0079] The sub-image generation error may include generating at least one sub-image that includes at least two adjacent regions.

[0080] At least two adjacent regions located within one sub-image may correspond to touching particles in the sample.

[0081] At least one or more portions along the boundary of one of the adjacent regions may be continuous with at least one or more portions along the boundary of another of the adjacent regions.

[0082] Correcting sub-image generation errors may involve processing each of the adjacent regions within one sub-image separately.

[0083] Additionally, the first segmentation component may be configured to process the sub-images individually if more than one sub-image is provided.

[0084] The first segmentation component can also be configured to determine a second scan location of the sub-image.

[0085] In particular, the first segmentation component may be configured to assign contiguous portions of corresponding regions in each sub-image to clusters using a k-means clustering algorithm.

[0086] Thus, each region of each sub-image may contain at least one or more clusters.

[0087] The first segmentation component may be further configured to apply a flood-fill algorithm, where the flood-fill algorithm may be configured to generate a mask for at least one of the clusters.

[0088] Generating a mask for at least one of the clusters may include assigning the same value of color and / or intensity to contiguous portions within the corresponding cluster.

[0089] Each mask may delimit a portion of the corresponding region.

[0090] Portions within each region of the at least one image may correspond to mineral grains within a corresponding grain of the sample.

[0091] The first segmentation component can be configured to determine one second scan location for each mask.

[0092] Each second scan location may correspond to the center of gravity of a respective mask.

[0093] Additionally, the first segmentation component can be configured to correct for over-segmentation errors.

[0094] The over-segmentation error may include determining more than one second scan location for at least one of the masks.

[0095] Thus, the first segmentation component may be configured to merge, via a merge operator, the second scan locations for at least one of the masks into one second scan location.

[0096] The data processing system may be configured to generate a first data set for a region within each sub-image.

[0097] The first data set may include a list of coordinates of second scan locations associated with one of the regions.

[0098] The data storage component may be configured to provide at least one of the first data sets.

[0099] The scanning microscope system may be configured to focus the beam onto at least one of the second scan locations of at least one of the first data sets.

[0100] The second detector may be configured to detect a corresponding second emission arising from a corresponding second scan location of the first data set when the sample is illuminated with the beam.

[0101] The first segmentation component may be configured to process at least one or more sub-images for the duration of a first time interval. The first time interval may correspond to the difference between an initial time and a final time. The initial time may correspond to a time at which the segmentation component determines a first one of the second scan locations of a first one of the sub-images. The final time may correspond to a time at which the segmentation component determines a last one of the second scan locations of a last one of the sub-images.

[0102] Further, the scanning microscope system may be configured to focus the beam on a second scan location of the corresponding first data set and detect the corresponding second emission for the duration of a second time interval. The second time interval may correspond to the difference between another initial time and another final time. The other initial time may correspond to the time at which the second detector detects the second emission from a first scan location of the second scan locations of a first data set of the first data set. The other final time may correspond to the time at which the second detector detects the second emission from a last scan location of the second scan locations of a last data set of the first data set (associated with the last sub-image).

[0103] The system may further be configured to, when two or more sub-images (i.e., first data sets) are provided, determine second scanning locations of the sub-images and detect second emissions from at least one of the second scanning locations of the first data set in parallel.

[0104] In other words, the first time interval and the second time interval may overlap. In this example, the system may be configured to (sequentially) detect a second emission from a second scan location within a first data set of one sub-image while simultaneously selecting another sub-image and (sequentially) generating another set of second scan locations. The system may be configured to repeat the process until all sub-images have been selected and processed. In general, parallel implementation of two method steps may correspond to partial or complete overlap of corresponding time intervals. The above considerations may apply to other subsequent parallel steps.

[0105] The data processing system may be configured to generate a first spectrum based on the second emission (i.e., the number of photons) detected at each of the second scanning locations of the at least one sub-image (i.e., the first data set).

[0106] The data processing system (800), and in particular the data storage component (810), may be configured to provide a first spectrum.

[0107] Each first spectrum may correspond to an X-ray spectrum, the X-ray spectrum including at least one or more spectral lines.

[0108] The X-ray spectrum may further include the number of X-ray photons detected at each energy (ie, the spectral line intensities).

[0109] Each spectral line may correspond to an electronic transition of a chemical element.

[0110] Each mineral grain of the sample may contain at least one or more chemical elements.

[0111] Thus, the X-ray spectrum may contain information regarding the chemical composition (eg, mineral composition) of the corresponding mineral grains associated with each second scan location.

[0112] The data processing system may further comprise a first spectral analysis component.

[0113] The data processing system, and in particular the first spectrum analysis component, may be configured to analyze each first spectrum from a respective second scan location of at least one of the regions (i.e., the first data set).

[0114] Analyzing each first spectrum may include comparing the respective first spectrum to at least one or more reference spectra.

[0115] Each reference spectrum may comprise a number of predetermined spectral lines associated with known mineral grains.

[0116] The first spectral analysis component may include a first line assignment component, and the first line assignment component may be configured to assign spectral lines of each first spectrum to predetermined spectral lines of the reference spectrum.

[0117] As such, the data processing system, and in particular the first spectrum analysis component, may be configured to match each first spectrum to one of the known mineral grains based on the results of the first line assignment component.

[0118] The data processing system, and in particular the first spectrum analysis component, may be configured to calculate a reliability score for every first spectrum, where the reliability score may correspond to a level of match between the first spectrum and the corresponding matched reference spectrum.

[0119] In other words, the confidence score may represent the probability that each first spectrum belongs to one of the known mineral grains.

[0120] In particular, a reliability score may correspond to a numerical value, which may range from 0 to a maximum of 1, that may be assigned to each first spectrum.

[0121] The system may be configured to pre-set thresholds for the reliability scores.

[0122] A high confidence score (above a threshold) may correspond to a reliable identification of the chemical composition of each mineral grain based on the corresponding first spectrum. For example, high confidence may correspond to a numerical value of 0.95 or greater. The confidence score may be expressed as a probability percentage, e.g., 95%.

[0123] On the other hand, a low confidence score (below the threshold) may correspond to a partial identification of the chemical composition of each mineral grain based on the corresponding first spectrum. For example, this may be the case when multiple spectral lines of the corresponding first spectrum can be assigned to two or more mineral grains (because their chemical compositions are similar). Other spectral lines that are unique to only one of the candidate mineral grains may have insufficient line intensity.

[0124] In other words, a mineral grain with a low reliability score may have at least a similar chemical composition to another mineral grain.

[0125] Thus, mineral grains of similar chemical composition may contain at least one or more common chemical elements.

[0126] Thus, mineral grains of similar chemical composition may correspond to similar first spectra, and indeed, similar first spectra may include at least one or more common spectral lines.

[0127] Furthermore, mineral grains of similar chemical composition may contain the same chemical elements in different elemental ratios. For example, the minerals Fe2O3 and Fe3O4, also known as hematite and magnetite, respectively, are both iron oxides but have different iron (and oxygen) contents. Therefore, corresponding first spectra may contain the same spectral lines (at the same energy) but with different intensity ratios (e.g., the ratio of iron line intensity to oxygen line intensity within each first spectrum).

[0128] As described above, a low reliability score may be due to poor spectral quality of each first spectrum, which may be due to spectral lines of each first spectrum having insufficient spectral line intensity, which may be due to an insufficient number of photons detected during the second dwell period (i.e., the number of photons detected per time).

[0129] The system, particularly the data processing system, may be configured to estimate an intrinsic photon count rate based on the number of photons detected during the second dwell period (i.e., the number of detected photons per time period).

[0130] The specific photon count rate may depend on the chemical composition of the corresponding part (i.e., mineral grain).

[0131] The intrinsic photon count rate may also depend on sample properties such as the crystalline characteristics (e.g., orientation, size, depth) of the corresponding part (i.e., mineral grain).

[0132] The system, in particular the data processing system, may be further configured to generate at least one or more second datasets, each second dataset comprising a list of coordinates of second scanning locations associated with mineral grains with low confidence scores in at least one or more regions (i.e., sub-images) of the at least one image.

[0133] The data processing system, and in particular the data storage component, may be configured to provide a second data set.

[0134] The system may be further configured to detect second emissions from second scan locations of the first data set and provide first spectra in parallel when more than second scan locations are processed. The first spectra may be provided in batches, and each batch of first spectra may result from second scan locations of one or more sub-images.

[0135] Additionally, the system may be configured to detect a second emission from a second scan location of the first data set and analyze the first spectra in parallel when two or more first spectra are provided.

[0136] The data processing system may include a one-pass classification component.

[0137] Additionally, the data processing system may include a two-pass classification component.

[0138] The system, particularly the data processing system, may be configured to perform a one-pass and / or two-pass classification component based on the results of the first spectral analysis component. Indeed, the system may be configured to apply a one-pass classification component to second scanned locations of the second data set associated with mineral grains that have a similar chemical composition as at least one other mineral grain. Furthermore, the system may be configured to apply a two-pass classification component to second scanned locations of the second data set associated with a particular mineral grain. Apart from similar composition, a particular mineral grain may further comprise a similar gray level intensity on at least one image as at least one other particular mineral grain (see below for further details).

[0139] In either case, the selected second scan location may refer to a mineral grain (and first spectrum) with a low confidence score. The system may also be configured to perform the one-pass and two-pass classification components, or portions thereof, in parallel. For example, the system may be configured to apply a one-pass classification component to one second scan location of a corresponding second data set and a two-pass classification component to another second scan location of the same second data set. Additionally or alternatively, the system may also be configured to apply a one-pass classification component to one second scan location of one second data set and a two-pass classification component to another second scan location of another second data set. Additionally or alternatively, the system may be configured to apply the two classification steps separately (i.e., at different, non-overlapping time intervals).

[0140] Further, the system, particularly the data processing system, may be configured to perform a one-pass and / or two-pass classification component after the second detector completes detection of the second emission from at least some or all of the second scan locations of the first data set of at least one image.

[0141] The system, in particular the scanning microscope system, may be configured to focus the beam onto at least one of the second scan locations of the second data set.

[0142] The scanning microscope system, and in particular the second detector, may be configured to detect a second emission from each second scan location of the second data set when the sample is illuminated with the beam.

[0143] The scanning microscope system, and in particular the second detector, may be configured to detect a second emission from at least one second scan location of the second data set for the duration of a third dwell period.

[0144] The data processing system, particularly the one-pass classification component, may be configured to determine a third dwell period for at least one second scan location of the second data set based on the calculated confidence score of each first spectrum and the characteristic photon count rate of the corresponding mineral grain. This may therefore optionally be advantageous because it may enable the system to dynamically adjust and optimize the acquisition of second emissions (e.g., X-rays) for each selected second scan location of the second data set.

[0145] In fact, the third dwell period may be longer than the second dwell period of the corresponding second scan location. For example, the third dwell period may be two to ten times longer than the second dwell period. Thus, the second detector may be configured to detect more photons (i.e., X-ray photons) within the third dwell period (relative to the second dwell period).

[0146] However, the third dwell period may also be less than or equal to the second dwell period of the corresponding second scan location.

[0147] The data processing system, and in particular the one-pass classification component, may be configured to add the number of photons (e.g., X-ray photons) detected within the second dwell period to the number of photons (e.g., X-ray photons) detected within the third dwell period at each second scan location of the second data set.

[0148] The data processing system, particularly the one-pass classification component (850a), may also be configured to generate at least one or more second spectra, each of which may include a total number of photons (e.g., X-ray photons) detected at a corresponding second scan location of the second data set. Thus, optionally and advantageously, the second spectra may have higher spectral intensity and / or resolution than the first spectrum for the corresponding selected second scan location.

[0149] The data processing system, particularly the data storage component, may be configured to provide the second spectra in groups, each group comprising second spectra obtained from at least one or more images. In particular, the system may be configured to acquire a grid of images, each grid comprising, for example, 8x8 images. Each image within a respective grid may depict a section of the sample, and thus the grid of images may depict adjacent sections of the sample. Thus, a group may comprise second X-ray spectra acquired from all images (8x8 images) within a corresponding grid.

[0150] Each second spectrum may correspond to an X-ray spectrum.

[0151] The data processing system, and in particular the one-pass classification component, may be configured to correct for image generation errors when a group comprises a second spectrum obtained from at least two images, and the at least two images may show adjacent sections of the sample.

[0152] The at least two images may include at least two portions belonging to one of the mineral grains with a low confidence score, in other words, a portion of one mineral grain with a low confidence score may be depicted in one of the two images, and another portion of the same mineral grain may be depicted in the other of the two images.

[0153] Correcting the imaging error may include stitching at least two portions of the at least two images.

[0154] Therefore, the data processing system, and in particular the one-pass classification component, may be configured to use another merge operator to sum the second spectra of parts belonging to the same mineral grain with a lower confidence score.

[0155] Additionally, the data processing system, and in particular the one-pass classification component, may include a second spectral analysis component.

[0156] The one-pass classification component, particularly the second spectrum analysis component, may be configured to analyze the second spectrum of each group separately.

[0157] In particular, the second spectral analysis component may be configured to match each second spectrum to at least two known mineral grains based on the results of the second line assignment component.

[0158] The data processing system, and in particular the second spectrum analysis component, may be configured to calculate at least two or more new confidence scores for each second spectrum, each new confidence score corresponding to a level of match between the second spectrum and a corresponding matched reference spectrum.

[0159] The one-pass classification component can be configured to select the highest new confidence score from among the at least two new confidence scores for each second spectrum.

[0160] The highest new reliability scores of at least some of the second spectra may correspond to a high reliability score (above the threshold). In other words, a statistical average of the highest new reliability scores of all the second spectra may correspond to a high reliability score.

[0161] Therefore, the data processing system, particularly the one-pass classification component, may be configured to normalize the data quality of at least some of the mineral grains with (initial) low confidence scores. For example, a mineral grain may be initially assigned to chalcocite (CuS) with 86% accuracy, while another mineral grain from the same sample may be assigned to chalcocite (CuS) with 74% accuracy. Once the one-pass classification process is complete, both minerals may be identified as chalcocite and chalcocite, respectively, with 99.9% accuracy.

[0162] Thus, optionally and advantageously, a one-pass classification process may allow for a reduction in overall acquisition time, as unambiguous mineral grains may be identified with a minimum of photons, while mineral grains with increasingly similar chemical compositions are selected for a second round of X-ray detection, whereby additional photons are acquired. As noted above, the acquisition of additional photons leads to a highly resolved second spectrum that may help accurately identify mineral grains with similar chemical compositions from one another.

[0163] In other words, the one-pass classification component requires approximately 2 x 10 per second scan location, as done in the previous slower methods of the prior art. 4 Instead of detecting photons, we average at least 2 x 10 per second scan location. 3 photons and up to 3×10 3 photons. Thus, optionally and advantageously, the one-pass sorting step may be several times and at least five times, preferably at least ten times, faster than conventional sorting processes.

[0164] The system may be further configured to detect second emissions from second scan locations of the second data set and analyze the first spectrum in parallel. In other words, once the system finishes detecting second emissions from second scan locations of the first data set, the system immediately begins a second round of detection (for second scan locations of the second data set) and finishes applying the first spectral analysis component to the remaining second scan locations of the first data set.

[0165] The system, particularly the data processing system, may be configured to execute the second spectral analysis component after the first spectral analysis component has completed analysis of at least some or all of the first spectra associated with at least one or more images.

[0166] Additionally, the system may be configured to detect a second emission from a second scan location of a second data set and analyze the second spectrum in parallel.

[0167] In other words, the system may be configured to analyze a first group of second spectra associated with a first image and to detect in parallel a second emission from a second scan location associated with a second image.

[0168] As described above, at least one or more mineral grains with a low confidence score associated with the second data set may correspond to a particular mineral grain, which may be depicted with the same or similar intensity (i.e., gray level intensity) on at least one image as at least one other particular mineral grain.

[0169] Thus, one particular mineral grain may be indistinguishable from at least one other particular mineral grain on at least one image. The particular mineral grain may have been determined so from a previous measurement of a replicant sample. The system may further include a list of particular mineral grains that require further processing. If the first spectrum matches one of the particular mineral grains on the list (with a low confidence score), the system may be notified that it may have an inadequately segmented portion of at least one image. An example of a pair of such particular mineral grains may include the mineral grains Fe2O3 and Fe3O4.

[0170] Furthermore, the portion of the selected second scan location associated with a particular mineral grain may correspond to a particular portion of the corresponding sub-image.

[0171] The system, particularly the data processing system, may be configured to perform a two-pass classification component on selected second scan locations of the second data set that are associated with particular portions of the corresponding sub-images.

[0172] Additionally, the system may be configured to preset and / or control the contrast and brightness values ​​of a corresponding image (e.g., at least one image or section thereof) by adjusting the operational settings of the first detector prior to detection of the corresponding first emission. Controlling the contrast and brightness values ​​may include changing the operational settings with respect to a set of default operational settings. The system may be configured to set the brightness value of at least one image to be equal to the midpoint or another estimate of the gray level intensity of a particular portion. The system may further be configured to increase the contrast to utilize the full pixel depth across variations in grain intensity. This step may correspond to an automated method step performed by a corresponding algorithm.

[0173] Adjusting the operating settings of the first detector may include adjusting a gain factor of at least one or more amplifiers integrated within the first detector.

[0174] Additionally or alternatively, adjusting the operational settings of the first detector may include adjusting a code width of an analog-to-digital converter (AD converter) integrated within the first detector.

[0175] The system, particularly the scanning microscope system, may be configured to focus the beam at a first scan location within a particular portion of the corresponding sub-image.

[0176] The scanning microscope system, and in particular the first detector, may be configured to redetect the first emission from the first scan location of the corresponding particular portion when the beam illuminates the sample.

[0177] The data processing system, and in particular the two-pass classification component, may be configured to generate at least one or more new sub-images, each new sub-image based on the first emission re-detected in each particular portion.

[0178] The system may be further configured to acquire new sub-images with adjusted contrast and brightness values ​​(for at least one image) by presetting the operational settings of the first detector accordingly (as described above). The adjusted contrast value may correspond to an increased contrast value. The adjusted contrast and brightness values ​​may vary for each particular sub-image.

[0179] The data processing system, and in particular the data storage component, may be configured to provide the new sub-images.

[0180] The new image may correspond to a new sub-image, and the portion bounded by the new image may correspond to a particular portion bounded by the new sub-image, such that each new sub-image includes one of the particular portions of each sub-image.

[0181] The data processing system, in particular the two-pass classification component, may be further configured to reveal and / or detect at least two or more new portions in at least one of the new sub-images using the adjusted contrast and brightness values ​​of each new sub-image.

[0182] Each new portion may correspond to a section of a particular portion within the respective new sub-image.

[0183] The data processing system, particularly the two-pass classification component, may be configured to identify at least two mineral grains in at least one new sub-image based on the different intensities (e.g., gray level intensities) of each new portion, where the two mineral grains are indistinguishable in at least one image (and sub-image), and one of the mineral grains may correspond to a specific mineral grain in the specific portion. For example, if a specific portion may initially be assigned to a specific mineral particle, Fe2O3, with a low confidence score, and the adjusted contrast and brightness values ​​of the new sub-image reveal two new portions, one new portion may be associated with Fe2O3 and the other new portion may be associated with Fe3O4. In other words, a specific portion may correspond to both iron oxide minerals, but only one appears to be included due to the low and / or suboptimal contrast and brightness values ​​of the respective sub-image (and image).

[0184] If no new moieties are revealed in one of the new sub-images, the two-pass classification component may be further configured to confirm that a particular moiety corresponds to only one mineral grain.

[0185] The two-pass classification component may further include a second segmentation component.

[0186] The two-pass classification component, and in particular the second segmentation component, may be configured to process new sub-images individually when more than one new sub-image is provided.

[0187] The two-pass classification component, and in particular the second segmentation component, may be configured to generate a new second scan location for each new portion in the at least one new sub-image using a k-means clustering algorithm and a flood-fill algorithm.

[0188] The two-pass classification component, particularly the second segmentation component, may be configured to determine each new second scan location for the duration of another segmentation period.

[0189] The data processing system, particularly the two-pass classification component, may be configured to generate third data sets, each of which may include a list of coordinates of new second scan locations associated with one of the new sub-images.

[0190] The data processing system, and in particular the data storage component, may be configured to provide at least one of the third data sets.

[0191] The system, in particular the scanning microscope system, may be configured to focus the beam onto at least one of the new second scan locations of the at least one third data set.

[0192] The scanning microscope system, and in particular the second detector, may be configured to detect a second emission from a new second scan location of a third data set as the sample is illuminated with the beam.

[0193] The scanning microscope system, and in particular the second detector, may be configured to detect the second emission for the duration of another third dwell period for each new second scan location of the third data set, which other third dwell periods may vary for each new second scan location within the same third data set and / or between different third data sets.

[0194] The system may also be configured, when two or more new sub-images are provided, to determine new second scan locations for the new sub-images and to detect second emissions from the new second scan locations in the third data set in parallel.

[0195] Additionally, the system may be configured to determine new second scan locations of the third data set and analyze the first spectra in parallel. In other words, once the system finishes detecting the second emission from the second scan location of the first data set, the system immediately begins acquiring and segmenting new sub-images (associated with the second scan location of the second data set) and finishes analyzing the first spectra of the remaining second scan locations of the first data set.

[0196] Thus, the system may also be configured to detect a second emission from a new second scan location of a third data set and analyze the first spectrum in parallel.

[0197] The data processing system, in particular the two-pass classification component, may be further configured to generate at least one or more new spectra, each new spectrum being based on the second emission detected at a respective new second scan location of the at least one third data set.

[0198] The data processing system, and in particular the data storage component, may be configured to provide new spectra to other groups, each other group including new spectra obtained from at least one or more images.

[0199] The data processing system, and in particular the two-pass classification component, may be configured to correct for image generation errors as done by the one-pass classification component.

[0200] The two-pass classification component may include another second spectral analysis component, which may include analyzing each other group of new spectra individually.

[0201] The two-pass classification component, particularly the other second spectral analysis component, may also be configured to compare each new spectrum with a reference spectrum and calculate at least two or more other new confidence scores, as performed by the second spectral analysis component.

[0202] The data processing system, and in particular the two-pass classification component, may be configured to execute another second spectral analysis component after the first spectral analysis component has completed analysis of at least some or all of the first spectra associated with at least one or more images.

[0203] The system may also be configured to detect a second emission from a new, second scan location of the third data set and to analyze in parallel a new spectrum associated with the new, second scan location.

[0204] Indeed, the system may be configured to analyze a first other group of new spectra associated with the first image and to detect in parallel a second emission from a new second scan location associated with the second image.

[0205] The system, particularly the two-pass classification component, may be configured to identify and classify at least two specific mineral grains within the same new sub-image based on the results of the second segmentation component and other second spectral analysis components. In other words, the gray level intensity may reveal, for example, which of the at least two mineral grains within the same new sub-image is the lightest (low average atomic number, e.g., Fe2O3) and which is the heaviest (high average atomic number, e.g., Fe3O4), since the intensity is related to the atomic number (e.g., average atomic number) of each mineral grain. Thus, the system may be configured to provide complementary information for each new spectrum, thereby providing a reliable classification of specific mineral grains within the corresponding new sub-image.

[0206] Thus, optionally and advantageously, the system may allow for automatic pre-selection of mineral grains of interest and further processing of them without reprocessing readily identifiable grains. As a result, the system may be advantageous because it can correctly distinguish between similar (i.e., specific) mineral grains within a sample in fully automated multimodal processing while maintaining high system throughput.

[0207] The third dwell period can be longer than the first dwell period.

[0208] The other third dwell period may be longer than the first dwell period.

[0209] The other segmentation periods may be longer than the other third dwell periods. However, the other segmentation periods may be shorter than the other third dwell periods. Similar to the segmentation periods of the first segmentation component, the other segmentation periods may also depend on the grain size and image characteristics.

[0210] Furthermore, operating the (first and second) segmentation components and the (first, second, other second) spectral analysis components (i.e., computational resources) in parallel with the scanning microscope system (i.e., first and second detectors, instrumentation resources) may optionally be advantageous, as this may allow for increased overall resource utilization of the system.

[0211] The data processing system may further include a post-processing component.

[0212] The post-processing component may be configured to obtain other images from other sections of the sample and / or replicant sample at localized contrast and brightness values ​​of at least one of the new sub-images.

[0213] The post-processing component may be configured to apply a calibration model, which may be configured to relate the chemical composition (e.g., average atomic number) of the mineral grains of the sample to the gray level intensity of corresponding portions of each other image.

[0214] The calibration model may correspond to a linear regression (Equation 1 above: I = C * See S+B).

[0215] Additionally, the calibration model may include contrast and brightness values ​​as model parameters (ie, fitting parameters).

[0216] Applying the calibration model may involve calibrating and further constraining localized contrast and brightness values ​​using known mineral grains (i.e., mineral elements) and their corresponding gray-level intensities in each of the other images. When using three or more known mineral grains (i.e., mineral elements), these values ​​may be determined with greater accuracy. Thus, a user may obtain a calibration curve for a sample of interest or a section thereof.

[0217] Thus, optionally and advantageously, the post-processing component may be configured to determine the chemical composition of unknown mineral grains of the sample and / or other sections of the replicant sample based on the corresponding calibration curves (i.e., based on their detected grey level intensities and calibrated contrast and brightness values).

[0218] The system may be a system configured for materials analysis and mineralogy.

[0219] The sample may include a plurality of particles embedded in an epoxy matrix.

[0220] The size of each mineral grain may include dimensions ranging from at least 1 μm to a maximum of 500 μm.

[0221] The scan points may include dimensions up to 1 micron.

[0222] At least one portion of the image (eg, a pixel) may include a dimension in the range of at least 10 nm to at most 1000 nm.

[0223] The second dwell period can range from at least 1 millisecond to at most 10 milliseconds.

[0224] The system may further include a control unit, which may be configured to control the power supply and operation of several components of the scanning microscope system, such as the condenser lens, the objective lens, the scanning coil, and the movable stage.

[0225] The system may also include a vacuum system, which may include a vacuum controller, a mechanical pumping system, an ultra-high vacuum pump, and a vacuum chamber.

[0226] The mechanical pumping system and the ultra-high vacuum pump may be configured to provide an ultra-high vacuum within the vacuum chamber.

[0227] The vacuum chamber may be configured to contain a sample, a movable stage, a first detector, a second detector or portions thereof, and a scanning electron microscope or portions thereof.

[0228] In a second embodiment, a method is disclosed, and the definitions, details, and advantages discussed above in the context of the system may each apply.

[0229] The method includes providing at least one or more images of a sample or section thereof based on first emissions detected from a plurality of first scan locations within a first dwell period. The method also includes performing a first detection step. The first detection step includes detecting second emissions from at least one or more second scan locations in at least one region of the at least one image during a second dwell period, each second scan location associated with a corresponding portion of the region. The method further includes performing a first spectrum providing step. The first spectrum providing step includes providing at least one or more first spectra, each first spectrum based on second emissions detected at each of the second scan locations in the at least one region. The method also includes performing a first spectrum analysis step. The first spectrum analysis step includes calculating a reliability score for all first spectra and selecting second scan locations associated with first spectra having reliability scores below a threshold. The method further includes performing a classification step. The classification step includes detecting a second emission from at least one of the selected second scan locations during a third dwell period and / or providing at least one or more new images delimiting portions associated with the selected second scan locations, and determining a new second scan location in the corresponding new image via the modified contrast and brightness values ​​thereof for the at least one image.

[0230] The method may further include a first segmentation step, the first segmentation step including determining a second scan location of the at least one region of the image.

[0231] The method may further include focusing a beam of charged particles (such as electrons) to a scan point on the sample.

[0232] The method may further include scanning the beam of charged particles across a plurality of scan locations in one or two dimensions.

[0233] The scan location may correspond to the first scan location.

[0234] The scan location may also correspond to a second scan location.

[0235] Additionally, the method may include assigning a two-dimensional coordinate system to the sample.

[0236] The method may also include assigning a two-dimensional coordinate system of the sample to at least one image.

[0237] The results of scanning a beam of charged particles over a scan location on a sample may include interactions between the beam and the sample.

[0238] Additionally, the result of the interaction may include a first and / or second emission.

[0239] The first emission may include an emission of particles (such as backscattered electrons).

[0240] The second emission may include the emission of photons (such as X-rays).

[0241] The method may further include detecting the first emission from each first scan location.

[0242] Additionally, the method may include generating at least one image based on the first emissions detected at each first scan location.

[0243] At least one image may correspond to a backscattered electron image.

[0244] Additionally, at least one image may show intensity variations between regions (or portions thereof) having different properties (such as chemical composition).

[0245] The intensity variation may include a gray level variation.

[0246] At least one image may include contrast and brightness values.

[0247] Each region of the at least one image may correspond to a particle in the sample.

[0248] Each particle in the sample may contain at least one or more mineral grains.

[0249] The method may further include detecting the first emission for the duration of the first dwell period at each first scan location.

[0250] The detecting step may further include detecting a second emission at each second scan location for the duration of a second dwell period.

[0251] The second dwell period may be longer than the first dwell period.

[0252] The method may further include detecting the first emission and detecting the second emission at different time intervals, wherein the different time intervals may correspond to non-overlapping time intervals.

[0253] The first segmentation step may include determining each second scan location for the duration of the segmentation period.

[0254] The segmentation period may depend on image characteristics such as the resolution and magnification of at least one image (and / or section thereof).

[0255] As mentioned above, the segmentation period may also depend on the size of the mineral grains and / or particles. However, due to the low statistical spread of sample statistics (e.g., grain size and number of grains per particle), the variation in segmentation period from one second scan location to another within the same sample and / or replicant samples may be small.

[0256] The segmentation period may be less than or equal to the second dwell period.

[0257] However, the segmentation period may be longer than the second dwell period, and indeed, large size mineral grains and / or particles associated with portions and / or regions of an image containing a large number of portions may lead to a long segmentation period (i.e., a slow segmentation process relative to the detection process).

[0258] The method may further comprise a pre-processing step.

[0259] The pre-processing step may include applying a thresholding algorithm.

[0260] The thresholding algorithm may include separating at least one image into background and foreground portions based on a threshold strength.

[0261] The background portion may include background portions (eg, pixels) that include intensities below a threshold intensity (eg, dark gray and / or black portions).

[0262] The foreground portion may include a foreground portion (eg, a pixel) that includes an intensity equal to or greater than a threshold intensity (eg, a light gray and / or white portion).

[0263] The thresholding algorithm may include segmenting / removing background portions, which may include assigning the same color and / or intensity value (e.g., black portions) to the background portions.

[0264] The foreground may include at least a portion of an area of ​​the image.

[0265] The pre-processing step may include determining the boundaries of the corresponding regions of the at least one image using a contour algorithm.

[0266] The contour algorithm may involve joining adjacent portions along the boundaries of corresponding regions with curves.

[0267] The adjacent portions along the border of the corresponding region may be surrounded by a background portion (eg, a black portion).

[0268] The pre-processing step may further include applying a bounding box algorithm, which may include dividing the at least one image into at least two or more sub-images based on the results of the contour algorithm.

[0269] A sub-image of at least one image may delimit a region.

[0270] The first segmentation step may include correcting sub-image generation errors.

[0271] The sub-image generation error may include generating at least one sub-image that includes at least two adjacent regions.

[0272] At least two adjacent regions located within one sub-image may correspond to touching particles in the sample.

[0273] At least one or more portions along the boundary of one of the adjacent regions may be continuous with at least one or more portions along the boundary of another of the adjacent regions.

[0274] Correcting sub-image generation errors may involve processing each of the adjacent regions within one sub-image separately.

[0275] The method may further include performing the first segmentation step separately on the at least two sub-images.

[0276] The first segmentation step may involve assigning contiguous portions of corresponding regions in each sub-image to clusters using a k-means clustering algorithm.

[0277] Thus, each region of each sub-image may contain at least one or more clusters.

[0278] The first segmentation step may further include applying a flood-fill algorithm, the flood-fill algorithm including generating a mask for at least one of the clusters.

[0279] Generating a mask for at least one of the clusters may include assigning the same value of color and / or intensity to contiguous portions within the corresponding cluster.

[0280] Each mask may delimit a portion of the corresponding region.

[0281] Portions within each region of the at least one image may correspond to mineral grains within a corresponding grain of the sample.

[0282] The first segmentation step may further include determining one second scan location for each mask.

[0283] Each second scan location may correspond to the center of gravity of a respective mask.

[0284] The method may further include generating a first data set for a region within each sub-image.

[0285] The first data set may include a list of coordinates of a second scan location associated with one of the regions.

[0286] The method may further include providing at least one of the first data sets.

[0287] The first detecting step may include focusing a beam onto at least one of the second scan locations of at least one of the first data sets.

[0288] The first detecting step may further include detecting a second emission from a corresponding second scan location of the first data set when the sample is irradiated with the beam.

[0289] Thus, the method may include performing the first segmentation step and the first detection step in parallel when two or more sub-images (i.e., first data sets) are provided.

[0290] The method may include generating a first spectrum based on second emissions (i.e., number of photons) detected at a corresponding second scanning location of at least one sub-image (i.e., first data set).

[0291] Each first spectrum may correspond to an X-ray spectrum.

[0292] An X-ray spectrum may include at least one or more spectral lines.

[0293] The X-ray spectrum may include the number of detected X-ray photons at each energy (ie, the spectral line intensities).

[0294] Each spectral line may correspond to an electronic transition of a chemical element.

[0295] Each mineral grain of the sample may contain at least one or more chemical elements.

[0296] The X-ray spectrum may include information regarding the chemical composition (eg, mineral composition) of the corresponding mineral grains associated with each second scan location.

[0297] The first spectral analysis step may include analyzing each first spectrum from each second scan location of the at least one region (ie, the first data set).

[0298] Analyzing each first spectrum may include comparing the respective first spectrum to at least one or more reference spectra.

[0299] Each reference spectrum may include a number of predetermined spectral lines associated with known mineral grains.

[0300] The first spectrum analysis step may further include a first line assignment step, which may include assigning spectral lines of each first spectrum to predetermined spectral lines of the reference spectrum.

[0301] The first spectral analysis step may include matching each first spectrum to one of the known mineral grains based on the results of the first line assignment step.

[0302] The confidence score may correspond to a level of match between the first spectrum and the corresponding matched reference spectrum.

[0303] Thus, the confidence score may represent the probability that each first spectrum belongs to one of the known mineral grains.

[0304] In practice, a reliability score may correspond to a numerical value, ranging from 0 to a maximum of 1, assigned to each first spectrum.

[0305] Additionally, the method may include presetting a threshold for the reliability score.

[0306] Thus, a high confidence score (above a threshold) may correspond to a reliable identification of the chemical composition of each mineral grain based on the corresponding first spectrum.

[0307] A low confidence score (below a threshold) may correspond to partial identification of the chemical composition of each mineral grain based on the corresponding first spectrum.

[0308] A mineral grain with a low confidence score may contain at least a similar chemical composition to another mineral grain.

[0309] Indeed, mineral grains of similar chemical composition may contain at least one or more chemical elements in common.

[0310] Furthermore, mineral grains of similar chemical composition may contain the same chemical elements in different elemental ratios (e.g., Fe2O3 and Fe3O4).

[0311] Thus, mineral grains of similar chemical composition may correspond to similar first spectra, and similar first spectra may include at least one or more common spectral lines.

[0312] Similar first spectra may contain the same spectral lines (same energy) at different intensity ratios.

[0313] The low reliability score may be due to the poor spectral quality of the respective first spectrum.

[0314] In particular, poor spectral quality may result from the spectral lines of each first spectrum containing insufficient spectral line intensity, which may result from an insufficient number of photons being detected during the second dwell period.

[0315] The method may further include estimating an intrinsic photon count rate based on the number of photons detected during the second dwell period (ie, the number of photons detected per time period).

[0316] The specific photon count rate may depend on the chemical composition of the corresponding part (i.e., mineral grain).

[0317] The intrinsic photon count rate may also depend on sample properties such as the crystalline characteristics (e.g., orientation, size, depth) of the corresponding part (i.e., mineral grain).

[0318] Additionally, the method may further include generating and providing at least one or more second data sets, each second data set may include a list of coordinates of second scan locations associated with mineral grains with low confidence scores in at least one or more regions (i.e., sub-images) of the at least one image.

[0319] The method may include performing the first detecting step and the first spectrum providing step in parallel.

[0320] The method may also include performing the first detecting step and the first spectral analyzing step in parallel.

[0321] The classification step may include a one-pass classification step and / or a two-pass classification step.

[0322] The classification step may include performing a one-pass and / or two-pass classification step based on the results of the first spectral analysis step.

[0323] Furthermore, the one-pass classification step may include performing a one-pass and / or two-pass classification step after a first detection step has essentially been performed on the corresponding image.

[0324] The one-pass classification step may include performing a one-pass and / or two-pass classification step after the first detection step is completed.

[0325] The one-pass classification step may include a second detection step, which may include focusing a beam onto at least one of the second scan locations of the second data set.

[0326] As described above, the second detecting step may further include detecting a second emission from each second scan location of the second data set as the sample is irradiated with the beam.

[0327] The second detecting step can include detecting a second emission from at least one second scan location (of the second data set) for the duration of a third dwell period.

[0328] The third dwell period may be longer than the second dwell period for the at least one second scan location.

[0329] The third dwell period may be less than or equal to the second dwell period for at least one second scan location.

[0330] The one-pass classification step may further include determining a third dwell period of at least one second scanning location of the second data set based on the calculated confidence score and the characteristic photon count rate of each first spectrum of the corresponding mineral grain.

[0331] The one-pass classification step may also include adding the number of photons (e.g., X-ray photons) detected by the first detection step to the number of photons (e.g., X-ray photons) detected by the second detection step at each second scanning location of the second data set.

[0332] The one-pass classification step may further include generating at least one or more second spectra, each of which may be based on second emissions detected by the first and second detection steps at respective second scanning locations of the second data set.

[0333] The method may include a second spectrum providing step, which may include providing the second spectra in groups, and each group may include second spectra obtained from at least one or more images.

[0334] Each second spectrum may correspond to an X-ray spectrum.

[0335] The one-pass classification step may include correcting for image generation errors where the group may include a second spectrum obtained from at least two images, where the at least two images may show adjacent sections of the sample.

[0336] The at least two images may include at least two portions belonging to one of the mineral particles with a low confidence score.

[0337] Correcting the imaging error may include stitching at least two portions of the at least two images.

[0338] The one-pass classification step may involve using another merge operator to sum a second spectrum of parts belonging to the same mineral grain with a low confidence score.

[0339] The one-pass classification step may include a second spectral analysis step.

[0340] The one-pass classification step may include performing a second spectral analysis step separately on the second spectra of each group.

[0341] The second spectral analysis step may include matching each second spectrum to at least two known mineral grains based on the results of the second line assignment step.

[0342] The second spectral analysis step may include calculating at least two or more new confidence scores for each second spectrum, and each new confidence score may correspond to a level of match between the second spectrum and a corresponding matched reference spectrum.

[0343] The one-pass classification step may include selecting the highest new confidence score from the at least two new confidence scores for each second spectrum.

[0344] The highest new confidence scores of at least some of the second spectra may correspond to high confidence scores (above a threshold).

[0345] The one-pass classification step may involve normalizing the data quality of mineral grains with (initial) low confidence scores.

[0346] The one-pass classification step may include performing the second detection step and the first spectral analysis step in parallel.

[0347] The one-pass classification step may include performing a first spectral analysis step essentially on at least one or more images followed by a second spectral analysis step.

[0348] The one-pass classification step may include performing a second detection step and a second spectral analysis step in parallel.

[0349] Additionally, the one-pass classification step may include performing a second spectral analysis step on a first group of second spectra associated with the first image and performing a second detection step in parallel on a second scan location associated with the second image.

[0350] At least one or more of the mineral grains with low confidence scores associated with the second data set may correspond to a particular mineral grain.

[0351] In particular, at least one of the particular mineral grains may be depicted on at least one image of a given contrast and brightness value with the same or similar intensity (i.e., gray level intensity) as at least one other particular mineral grain (e.g., Fe2O3 and Fe3O4).

[0352] Thus, at least one particular mineral grain may be indistinguishable from at least one other particular mineral grain on at least one image.

[0353] Furthermore, the portion of the selected second scan location associated with a particular mineral grain may correspond to a particular portion of the corresponding sub-image.

[0354] The classifying step may further include performing a two-pass classification step on selected second scanned locations of the second data set associated with the particular portion.

[0355] The two-pass classification step may include presetting and / or controlling the contrast and brightness values ​​of the corresponding image (e.g., at least one image or section thereof) by adjusting operational settings of at least one or more system components prior to detecting the corresponding first emission.

[0356] The two-pass classification step may further include focusing the beam at a first scan location within the particular portion.

[0357] Thus, the two-pass classification step may involve re-detecting the first emission from the first scan location of the corresponding particular portion as the beam illuminates the sample.

[0358] The method may include generating and providing at least one or more new sub-images, each new sub-image based on the re-detected first emission in each particular portion.

[0359] The two-pass classification step may further include acquiring a new sub-image with adjusted contrast and brightness values ​​by pre-adjusting the operating settings of at least one system component accordingly.

[0360] The new image may correspond to a new sub-image.

[0361] The portion bounded by the new image may correspond to the particular portion bounded by the new sub-images, and thus, as described above, each new sub-image may include one of the particular portions of the corresponding sub-image.

[0362] The two-pass classification step may further include using the adjusted contrast and brightness values ​​of each new sub-image to reveal and / or detect at least two or more new portions in at least one of the new sub-images.

[0363] Each new portion may correspond to a section of a particular portion within the respective new sub-image.

[0364] The two-pass classification step may include identifying at least two mineral grains in at least one new sub-image based on different intensities (e.g., gray level intensities) between each new portion, and the at least two mineral grains may be indistinguishable on at least one image (and sub-image), and one of the mineral grains may correspond to a particular mineral in a particular portion.

[0365] The two-pass classification step may further include, if two or more new sub-images are provided, performing a second segmentation step on the new sub-images individually.

[0366] The second segmentation step may include generating new second scan locations for each new portion in the at least one new sub-image using a k-means clustering algorithm and a flood-fill algorithm.

[0367] Additionally, the second segmentation step may include determining each new second scan location for the duration of another segmentation period.

[0368] The two-pass classification step may include generating third data sets, each of which may include a list of coordinates of new second scan locations associated with one of the new sub-images.

[0369] The method may further include providing at least one of the third data sets.

[0370] The two-pass classification step may include performing another second detection step on a new second scan location of the at least one third data set.

[0371] The other second detecting step may include focusing the beam onto at least one new second scan location of at least one of the third data sets.

[0372] Another second detection step may further include detecting a second emission from a corresponding new second scan location when the sample is irradiated with the beam.

[0373] Thus, another second detection step may include detecting a second emission for the duration of another third dwell period at each new second scan location of each third data set.

[0374] The two-pass classification step may include performing a second segmentation step and another second detection step in parallel when two or more new sub-images are provided.

[0375] Additionally, the two-pass classification step may include performing the second segmentation step and the first spectral analysis in parallel.

[0376] The two-pass classification step may also include performing another second detection step and a first spectral analysis in parallel.

[0377] The two-pass classification step may further include generating at least one or more new spectra, each new spectrum being based on second emissions detected by another second detection step at a respective new second scanning location of the at least one third data set.

[0378] Indeed, the method may include a step of providing new spectra, which may include providing new spectra to other groups, each of which may include new spectra obtained from at least one or more images.

[0379] The two-pass classification step may include correcting for image generation errors as done in the one-pass classification step.

[0380] The two-pass classification step may include performing a separate second spectral analysis step on each other group of new spectra individually.

[0381] Another second spectral analysis step may include comparing each new spectrum to a reference spectrum and calculating at least two or more other new confidence scores, as is done in the second spectral analysis step of the one-pass classification step.

[0382] The two-pass classification step may include essentially performing a first spectral analysis on at least one or more images followed by performing another second spectral analysis step.

[0383] The two-pass classification step may include performing another second detection step and another second spectral analysis step in parallel.

[0384] The two-pass classification step may include performing another second spectral analysis on a first other group of new spectra associated with the first image and performing a second detection step in parallel at a new second scanning location associated with the second image.

[0385] The two-pass classification step may include identifying and classifying at least two specific mineral grains within the same new sub-image based on the results of the second segmentation step and another second spectral analysis step.

[0386] The third dwell period may be longer than the first dwell period.

[0387] The other third dwell period may be longer than the first dwell period.

[0388] The other segmentation period may be longer than the other third dwell period. However, the other segmentation period may be shorter than the other third dwell period. As noted above, the other segmentation period may depend on the grain size and image characteristics.

[0389] The method may further include a post-processing step.

[0390] The post-processing step may include obtaining other images from other sections of the sample and / or replicant sample at localized contrast and brightness values ​​of at least one of the new sub-images.

[0391] The post-processing step may include applying a calibration model that relates the chemical composition (e.g., average atomic number) of the mineral grains of the sample to the gray level intensity of the corresponding portion on the at least one image.

[0392] The calibration model may correspond to a linear regression.

[0393] The calibration model may include contrast and brightness values ​​as model parameters (ie, fitting parameters).

[0394] Additionally, applying the calibration model may involve calibrating and further constraining localized contrast and brightness values ​​using mineral grains of known chemical composition and their corresponding gray level intensities in each other image.

[0395] Post-processing steps may include using the detected grey level intensities and calibrated contrast and brightness values ​​to determine the chemical composition of mineral grains associated with other sections of the sample and / or replicant sample.

[0396] At least a portion of the one-pass classification step and at least a portion of the two-pass classification step are computer-implemented.

[0397] Furthermore, the pre-processing step, the first segmentation step, the first spectral analysis step, and the post-processing step may be computer-implemented.

[0398] The method may be a method for materials analysis and mineralogy.

[0399] The sample may include a plurality of particles embedded in an epoxy matrix.

[0400] The size of each mineral grain may include dimensions ranging from at least 1 μm to a maximum of 500 μm.

[0401] The scan points may include dimensions up to 1 micron.

[0402] At least one portion of the image (eg, a pixel) may include a dimension in the range of at least 10 nm to at most 1000 nm.

[0403] The second dwell period can range from at least 1 millisecond to at most 10 milliseconds.

[0404] A single pass classification step averages at least 2x10 per second scan location. 3 photons and up to 3x10 3 The method may include calculating a high confidence score for at least some or all of the mineral grains in the sample by detecting the photons.

[0405] The method may include using a system according to any of the system embodiments.

[0406] The system may be configured to perform a method according to any of the aforementioned method embodiments.

[0407] In a third embodiment, a computer program product is disclosed.

[0408] The computer program product may include instructions that, when the program is executed by a computer, cause the computer to perform the steps of the methods disclosed above.

[0409] Another computer program product may include instructions that, when executed by the data processing system (800), cause the data processing system (800) to perform steps, the data processing system being configured for the steps.

[0410] The following embodiments also form part of the invention.

[0411] System Embodiments In the following, system embodiments are discussed. System embodiments are abbreviated by the letter "S" followed by a number. These embodiments are meant whenever reference is made herein to "system embodiments." S1. A system comprising a scanning microscope system (100) and a data processing system (800), the system comprising: - providing at least one or more images of the sample or sections thereof based on first emissions detected within a first dwell period from a plurality of first scanning locations; - detecting second emissions during a second dwell period from at least one or more second scan locations of at least one region of at least one image, each second scan location associated with a portion of the corresponding region; - providing at least one or more first spectra, each first spectrum being based on second emissions detected at each of second scan locations of the at least one region; - calculating a reliability score for all first spectra and selecting second scan locations associated with first spectra whose reliability scores are below a threshold; - detecting second emissions from at least one of the selected second scan locations during a third dwell period and / or providing at least one or more new images delimiting portions associated with the selected second scan locations, and determining a new second scan location in each new image via the modified contrast and brightness values ​​for the at least one image. S2. A system according to the previous embodiment, wherein the scanning microscope system (100) includes a first detector (111), the first detector configured to detect a first emission from a first scanning location. S3. A system according to any preceding embodiment, wherein the first detector includes a backscattered electron detector. S4. A system according to any of the previous embodiments, wherein the scanning microscope system (100) includes a second detector (112), the second detector configured to detect a second emission from the second scan location and the new second scan location. S5. The system according to any preceding embodiment, wherein the second detector includes an X-ray detector. S6. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is configured to focus a beam of charged particles (such as electrons) to a scanning point on a sample. S7. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is further configured to scan the beam of charged particles across a plurality of scan locations in one or two dimensions. S8. A system according to any of the previous embodiments, wherein the scan location corresponds to the first scan location. S9. A system according to any of the previous embodiments, wherein the scan location corresponds to a second scan location. S10. A system according to any of the previous embodiments, wherein the system, in particular the data processing system (800), is further configured to assign a two-dimensional coordinate system to the sample. S11. A system according to any of the previous embodiments, comprising the features of S10, wherein the system, in particular the data processing system (800), is configured to assign a two-dimensional coordinate system of the sample to at least one image. S12. A system according to any of the preceding embodiments, comprising the features of S7, wherein the result of scanning the beam of charged particles across the scanned location of the sample includes an interaction of the beam with the sample. S13. A system according to any preceding embodiment, comprising features of S1, wherein the interaction result includes a first and / or second emission. S14. A system according to any preceding embodiment, comprising features of S1 and S13, wherein the first emission comprises an emission of particles (such as backscattered electrons). S15. A system according to any preceding embodiment, comprising features of S1 and S13, wherein the second emission comprises an emission of photons (such as X-ray photons). S16. A system according to any of the previous embodiments, wherein the system, in particular the data processing system (800), is configured to generate at least one image based on the first emissions detected at each first scan location. S17. A system according to any of the previous embodiments, wherein at least one image corresponds to a backscattered electron image. S18. A system according to any of the preceding embodiments, wherein at least one image shows intensity variations between regions (and / or portions thereof) having different properties (such as chemical composition). S19. A system according to any preceding embodiment, wherein the intensity variation includes a gray level variation. S20. The system according to any of the previous embodiments, wherein at least one image includes contrast and brightness values. S21. A system according to any of the previous embodiments, wherein each region of at least one image corresponds to a particle in the sample. S22. A system according to any preceding embodiment, wherein each particle in the sample comprises at least one or more mineral grains. S23. A system according to any of the previous embodiments, wherein the scanning microscope system (100), in particular the first detector (111), is configured to detect the first emission for the duration of the first dwell period at each first scan location. S24. A system according to any of the previous embodiments, wherein the scanning microscope system (100), in particular the second detector (112), is configured to detect the second emission for the duration of the second dwell period at each second scan location. S25. A system according to either of the previous two embodiments, characterized by S1, wherein the second dwell period is longer than the first dwell period. S26. A system according to any of the preceding embodiments, wherein the system is configured to detect a first emission from a first scanning location and detect a second emission from a second scanning location at different time intervals, the different time intervals corresponding to non-overlapping time intervals. S27. A system according to any of the preceding embodiments, wherein the data processing system (800) includes a data storage component (810). S28. A system according to any of the preceding embodiments, wherein the data storage component (810) is configured to provide at least one image of the sample (or a section thereof). S29. A system according to any of the preceding embodiments, wherein the data processing system (800) includes a first segmentation component (820), and the first segmentation component (820) is configured to determine a second scan location of a region of at least one image. S30. A system according to any of the previous embodiments, wherein the data processing system, in particular the first segmentation component (820), is configured to determine each second scan location for the duration of the segmentation period. S31. A system according to any of the previous embodiments, wherein the segmentation period is longer than the second dwell period. S32. A system according to any of the preceding embodiments, wherein the segmentation period is less than or equal to the second dwell period. S33. A system according to any of the previous embodiments, wherein the segmentation period depends on image characteristics such as resolution and magnification of at least one image (and / or section thereof). S34. The system according to any of the previous embodiments, wherein the segmentation period depends on the size of the mineral grains and / or particles. S35. The system of any of the preceding method embodiments, wherein the data processing system includes a pre-processing component (830). S36. A system according to the previous embodiment, wherein the data processing system (800), in particular the pre-processing component (830), is configured to apply a thresholding algorithm. S37. A system according to the previous embodiment, wherein the pre-processing component (830), in particular the thresholding algorithm, is configured to separate at least one image into background and foreground portions based on a threshold strength. S38. The system according to any of the previous embodiments, wherein the background portion includes a background portion, and the background portion (e.g., pixels) includes an intensity lower than a threshold intensity (e.g., dark gray and / or black portions). S39. The system according to any of the preceding embodiments, wherein the foreground portion includes a foreground portion, and the foreground portion (e.g., pixel) includes an intensity equal to or greater than a threshold intensity (e.g., light gray and / or white portion). S40. A system according to any of the preceding embodiments, wherein segmenting / removing background portions includes assigning the background portions a color and / or intensity of the same value (e.g., black portions). S41. A system according to any of the previous embodiments, wherein the foreground portion comprises at least a portion of an area of ​​the image. S42. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the pre-processing component (830), is configured to determine boundaries of corresponding regions of at least one image using a contour algorithm. S43. A system according to any of the previous embodiments, wherein the pre-processing component (830), in particular the contour algorithm, is configured to join adjacent portions along the boundaries of the corresponding regions into curves. S44. A system according to the previous embodiment, wherein adjacent portions along the boundary of the corresponding region are surrounded by background portions (e.g., black portions). S45. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the pre-processing component (830), is further configured to apply a bounding box algorithm. S46. A system according to the previous embodiment, wherein the pre-processing component (830), in particular the bounding box algorithm, is configured to divide at least one image into at least one or more sub-images based on the results of the contour algorithm. S47. The system according to any of the preceding embodiments, comprising the feature of S46, wherein a sub-image of at least one image delimits a region. S48. A system according to any of the preceding embodiments, comprising the features of S46, wherein the data processing system (800), and in particular the data storage component (810), is configured to provide a sub-image. S49. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is configured to correct sub-image generation errors. S50. A system according to any preceding embodiment, wherein the sub-image generation error includes generating at least one sub-image that includes at least two adjacent regions. S51. A system according to any of the preceding embodiments, wherein at least two adjacent regions located within one sub-image correspond to touching particles within the sample. S52. A system according to any of the preceding embodiments, wherein at least one or more portions along a boundary of one of the adjacent regions is contiguous with at least one or more portions along a boundary of another of the adjacent regions. S53. A system according to any of the preceding embodiments, comprising the features of S49, wherein correcting sub-image generation errors includes processing each of adjacent regions within one sub-image separately. S54. A system according to any of the preceding embodiments, comprising the features of S46, wherein the data processing system (800), in particular the first segmentation component (820), is configured to process the sub-images individually if more than one sub-image is provided. S55. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is configured to determine a second scanning location of the sub-image. S56. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is configured to assign contiguous portions of portions of corresponding regions in each sub-image to clusters using a k-means clustering algorithm. S57. A system according to either of the previous two embodiments, wherein each region of each sub-image includes at least one cluster. S58. A system according to any of the preceding embodiments, comprising the features of S56, wherein the data processing system (800), in particular the first segmentation component (820), is further configured to apply a flood-fill algorithm, the flood-fill algorithm being configured to generate a mask for at least one of the clusters. S59. The system according to the preceding embodiment, wherein generating a mask for at least one of the clusters includes assigning the same value of color and / or intensity to contiguous portions in corresponding clusters. S60. A system according to any of the previous embodiments, wherein each mask delimits a portion of a corresponding region. S61. A system according to any preceding embodiment, wherein portions within each region of the at least one image correspond to mineral grains within a corresponding particle of the sample. S62. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is configured to determine one second scan location for each mask. S63. A system according to any of the previous embodiments, wherein each second scan location corresponds to a centroid of a respective mask. S64. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is further configured to correct over-segmentation errors. S65. A system according to any preceding embodiment, wherein the over-segmentation error includes determining two or more second scan locations for at least one of the masks. S66. A system according to either of the two previous embodiments, wherein the data processing system (800), in particular the first segmentation component (820), is configured to merge second scan locations for at least one of the masks into one second scan location by a merge operator. S67. A system according to any of the previous embodiments, wherein the system, in particular the data processing system (800), is configured to generate a first data set for a region within each sub-image. S68. A system according to any of the preceding embodiments, wherein the first data set includes a list of coordinates of second scan locations associated with one of the regions. S69. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the data storage component (810), is configured to provide at least one of the first data sets. S70. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is configured to focus the beam onto at least one of the second scan locations of at least one of the first data sets. S71. A scanning microscope system (100), in particular a system according to any of the previous embodiments, wherein the second detector (112) is configured to detect a second emission from a corresponding second scan location of the first data set when the sample is irradiated with the beam. S72. A system according to any of the preceding embodiments, wherein the system is further configured to, if two or more sub-images (i.e., first data sets) are provided, determine second scan locations of the sub-images and detect second emissions in parallel from at least one of the second scan locations of the first data set. S73. A system according to any of the previous embodiments, comprising features of S1, wherein the system, in particular the data processing system (800), is configured to generate a first spectrum based on the second emission (i.e., number of photons) detected at each of the second scan locations of the at least one sub-image (i.e., first data set). S74. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the data storage component (810), is configured to provide a first spectrum. S75. The system according to any of the preceding embodiments, wherein each first spectrum corresponds to an X-ray spectrum. S76. A system according to any preceding embodiment, wherein the X-ray spectrum comprises at least one or more spectral lines. S77. A system according to any of the preceding embodiments, comprising features of S75 and S76, wherein the X-ray spectrum comprises the number of detected X-ray photons at each energy (i.e., spectral line intensities). S78. A system according to any preceding embodiment, wherein each spectral line corresponds to an electronic transition of a chemical element. S79. A system according to any of the preceding embodiments, wherein each mineral grain of the sample comprises at least one or more chemical elements. S80. A system according to any of the preceding embodiments, wherein the X-ray spectrum includes information regarding the chemical composition (e.g., mineral composition) of corresponding mineral grains associated with each second scan location. S81. A system according to any of the previous embodiments, wherein the data processing system (800) further includes a first spectral analysis component (840). S82. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the first spectrum analysis component (840), is configured to analyze each first spectrum from a respective second scan location of at least one of the regions (i.e., the first data set). S83. A system according to any preceding embodiment, wherein analyzing each first spectrum includes comparing each first spectrum to at least one or more reference spectra. S84. A system according to any preceding embodiment, wherein each reference spectrum includes a plurality of predetermined spectral lines associated with known mineral grains. S85. A system according to any of the preceding embodiments, wherein the first spectral analysis component includes a first line assignment component, the first line assignment component configured to assign spectral lines of each first spectrum to predetermined spectral lines of the reference spectrum. S86. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the first spectrum analysis component (840), is configured to match each first spectrum to one of the known mineral grains based on the results of the first line assignment component. S87. A system according to any of the preceding embodiments, comprising the features of S1, wherein the data processing system (800), in particular the first spectrum analysis component (840), is configured to calculate a confidence score for every first spectrum, the confidence score corresponding to a level of match between the first spectrum and the corresponding matched reference spectrum. S88. The system according to any of the preceding embodiments, wherein the classification confidence score represents the probability that each first spectrum belongs to one of the known mineral grains. S89. The system according to any of the preceding embodiments, wherein a reliability score corresponds to a numerical value, the numerical value ranging from 0 to at most 1, assigned to each first spectrum. S90. A system according to any of the preceding embodiments, comprising the features of S1, wherein the system is configured to pre-set a threshold for the reliability score. S91. The system according to any of the preceding embodiments, comprising the features of S90, wherein a high confidence score (above a threshold) corresponds to a reliable identification of the chemical composition of each mineral grain based on the corresponding first spectrum. S92. The system according to any of the preceding embodiments, comprising the features of S90, wherein a low confidence score (below a threshold) corresponds to a partial identification of the chemical composition of each mineral grain based on the corresponding first spectrum. S93. A system according to any of the preceding embodiments, wherein a mineral grain with a low confidence score comprises a similar chemical composition as at least another mineral grain. S94. A system according to any preceding embodiment, wherein mineral grains of similar chemical composition contain at least one or more common chemical elements. S95. A system according to any of the preceding embodiments, wherein mineral grains of similar chemical composition contain the same chemical elements in different elemental ratios (e.g., Fe2O3 and Fe3O4). S96. The system according to any of the preceding embodiments, comprising the feature of S94, wherein mineral grains of similar chemical composition correspond to similar first spectra, and the similar first spectra include at least one or more common spectral lines. S97. A system according to any of the preceding embodiments, comprising the features of S95, wherein the similar first spectra include the same spectral lines (at the same energy) with different intensity ratios. S98. The system according to any of the preceding embodiments, comprising the features of S92, wherein the low confidence score is due to low spectral quality of the respective first spectrum. S99. A system according to any of the preceding embodiments, wherein the poor spectral quality results from spectral lines of each first spectrum having insufficient spectral line intensity, and the insufficient line intensity results from an insufficient number of photons detected during the second dwell period (i.e., the number of detected photons per time). S100. A system according to any of the preceding embodiments, wherein the system, in particular the data processing system (800), is configured to estimate an intrinsic photon count rate based on the number of photons detected during the second dwell period (i.e., the number of detected photons per time). S101. A system according to any of the preceding embodiments, characterized by S100, wherein the characteristic photon count rate depends on the chemical composition of the corresponding part (i.e., mineral grain). S102. A system according to any of the preceding embodiments, characterized by S100, wherein the characteristic photon count rate depends on sample properties such as crystalline properties (e.g., orientation, size, depth) of the corresponding portion (i.e., mineral grain). S103. A system according to any of the preceding embodiments, wherein the system, in particular the data processing system (800), is further configured to generate at least one or more second data sets, each second data set comprising a list of coordinates of second scanning locations associated with mineral grains with low confidence scores in at least one or more regions (i.e., sub-images) of at least one image. S104. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the data storage component (810), is configured to provide a second data set. S105. A system according to any of the preceding embodiments, wherein the system is further configured to detect a second emission from a second scan location of the first data set and provide a first spectrum in parallel. S106. A system according to any of the preceding embodiments, wherein the system is further configured to detect a second emission from a second scan location of the first data set and analyze the first spectrum in parallel. S107. A system according to any of the previous embodiments, wherein the data processing system (800) includes a one-pass classification component (850a). S108. A system according to any of the previous embodiments, wherein the data processing system (800) includes a two-pass classification component (850b). S109. A system according to any of the previous embodiments, wherein the system, in particular the data processing system (800), is configured to perform a one-pass and / or two-pass classification component based on the results of the first spectral analysis component (840). S110. A system according to any of the preceding embodiments, wherein the system, in particular the data processing system (800), is configured to perform a one-pass and / or two-pass classification component after the second detector (112) completes detection of second emissions from at least some or all second scan locations of the first data set of at least one image. S111. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is configured to focus the beam onto at least one of the second scan locations of the second data set. S112. A system according to the previous embodiment, wherein the scanning microscope system (100), in particular the second detector (112), is configured to detect a second emission from each second scan location of the second data set when the beam is irradiated onto the sample. S113. A system according to any of the preceding embodiments, comprising features of S1, wherein the scanning microscope system (100), in particular the second detector (112), is configured to detect a second emission for the duration of a third dwell period from at least one second scan location of the second data set. S114. A system according to any of the preceding embodiments, wherein for at least one second scan location, the third dwell period is longer than the second dwell period. S115. A system according to any of the preceding embodiments, wherein for at least one second scan location, the third dwell period is less than or equal to the second dwell period. S116. A system according to any of the preceding embodiments, comprising the features of S100, wherein the data processing system (800), in particular the one-pass classification component (850a), is configured to determine a third dwell period for at least one second scanning location of the second data set based on the calculated confidence score of each first spectrum and the characteristic photon count rate of the corresponding mineral grain. S117. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the one-pass classification component (850a), is configured to add the number of photons (e.g., X-ray photons) detected within the second dwell period to the number of photons (e.g., X-ray photons) detected within the third dwell period at each second scan location of the second data set. S118. A system according to any of the preceding embodiments, comprising the features of S117, wherein the data processing system (800), in particular the one-pass classification component (850a), is further configured to generate at least one or more second spectra, each second spectrum including a total number of photons (e.g., X-ray photons) detected at a corresponding second scan location of the second data set. S119. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the data storage component (810), is configured to provide the second spectra in groups, each group including second spectra obtained from at least one or more images. S120. The system according to any of the previous embodiments, wherein each second spectrum corresponds to an X-ray spectrum. S121. A system according to any of the preceding embodiments, comprising the features of S119, wherein the data processing system (800), in particular the one-pass classification component (850a), is configured to correct image generation errors when the group includes a second spectrum obtained from at least two images, the at least two images showing adjacent sections of the sample. S122. A system according to any of the preceding embodiments, wherein at least two images include at least two portions belonging to one of the mineral grains with a low confidence score. S123. The system according to any of the preceding embodiments, comprising the features of S121, wherein correcting image generation errors includes stitching at least two portions of at least two images. S124. A system according to any of the preceding embodiments, comprising the features of S123, wherein the data processing system (800), in particular the one-pass classification component (850a), is configured to sum, using another merging operator, second spectra of portions belonging to the same mineral grain with a lower confidence score. S125. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the one-pass classification component (850a), includes a second spectral analysis component. S126.1 A system according to any of the previous embodiments, wherein the path classification component (850a), in particular the second spectrum analysis component, is configured to analyze the second spectrum of each group separately. S127. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the second spectrum analysis component, is configured to match each second spectrum to at least two known mineral grains based on the results of the second line assignment component. S128. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the second spectrum analysis component, is configured to calculate at least two or more new confidence scores for each second spectrum, each new confidence score corresponding to a level of match between the second spectrum and a corresponding matched reference spectrum. S129.1 The system according to any of the preceding embodiments, comprising the feature of S128, wherein the path classification component is configured to select a highest new confidence score from among the at least two new confidence scores for each second spectrum. S130. The system according to any of the preceding embodiments, comprising the feature of S129, wherein the highest new confidence scores of at least some of the second spectra correspond to high confidence scores (above a threshold). S131. A system according to any of the previous embodiments, comprising features of S130, wherein the data processing system (800), in particular the one-pass classification component (850a), is configured to normalize the data quality of at least some of the mineral grains with (initial) low confidence scores. S132. A system according to any of the preceding embodiments, wherein the system is configured to detect a second emission from a second scan location of a second data set and analyze the first spectrum in parallel. S133. A system according to any of the preceding embodiments, wherein the system, in particular the data processing system (800), is configured to execute a second spectral analysis component after the first spectral analysis component has completed analysis of at least some or all of the first spectra associated with at least one or more images. S134. A system according to any of the preceding embodiments, comprising features of S132 and S133, wherein the system is configured to detect a second emission from a second scan location of a second data set and analyze the second spectrum in parallel. S135. A system according to any of the preceding embodiments, wherein the system is configured to analyze a first group of second spectra associated with the first image and detect in parallel a second emission from a second scan location associated with the second image. S136. The system according to any of the preceding embodiments, wherein at least one or more of the mineral grains with low confidence scores associated with the second dataset corresponds to a specific mineral grain. S137. A system according to any of the preceding embodiments, wherein at least one of the specific mineral grains is depicted at the same intensity (i.e., gray level intensity) on at least one image as at least one other specific mineral grain (e.g., Fe2O3 and Fe3O4). S138. A system according to any of the preceding embodiments, wherein at least one particular mineral grain is indistinguishable from at least one other particular mineral grain on at least one image. S139. A system according to any of the preceding embodiments, wherein the portion of the selected second scan location associated with a particular mineral grain corresponds to a particular portion of the respective sub-image. S140. A system according to any of the preceding embodiments, wherein the system, in particular the data processing system (800), is configured to perform a two-pass classification component (850b) on selected second scan locations of the second data set associated with particular portions of the corresponding sub-image. S141. A system according to any of the preceding embodiments, wherein the system is configured to preset and / or control contrast and brightness values ​​of a corresponding image (e.g., at least one image or section thereof) by adjusting operational settings of the first detector (111) prior to detection of the corresponding first emission. S142. The system according to any of the previous embodiments, wherein adjusting the operational settings of the first detector (111) includes adjusting a gain factor of at least one or more amplifiers integrated within the first detector (111). S143. The system according to any of the previous embodiments, wherein adjusting the operational settings of the first detector (111) includes adjusting a code width of an analog-to-digital converter (AD converter) integrated within the first detector (111). S144. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is configured to focus the beam at a first scan location within a particular portion. S145. A system according to any of the previous embodiments, wherein the scanning microscope system (100), in particular the first detector (111), is configured to redetect the first emission from the first scanning location of the corresponding specific portion. S146. A system according to any of the preceding embodiments, comprising the features of S145, wherein the data processing system (800), in particular the two-pass classification component (850b), is configured to generate at least one or more new sub-images, each new sub-image based on the re-detected first emission in each particular portion. S147. A system according to any of the previous embodiments comprising features S1 and S141-S143, wherein the system is further configured to acquire new sub-images with adjusted contrast and brightness values ​​by pre-adjusting the operating settings of the first detector (111) accordingly. S148. A system according to any of the previous embodiments, wherein the data processing system (800), and in particular the data storage component (810), is configured to provide a new sub-image. S149. A system according to any of the preceding embodiments, comprising the features of S1, wherein the new image corresponds to a new sub-image. S150. A system according to any of the preceding embodiments, comprising the features of S1, wherein the portion bounded by the new image corresponds to the particular portion bounded by the new sub-image. S151. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the two-pass classification component (850b), is further configured to reveal and / or detect at least two or more new portions in at least one of the new sub-images using the adjusted contrast and brightness values ​​of each new sub-image. S152. A method according to any of the preceding embodiments, characterized by S151, wherein each new portion corresponds to a section of a particular portion within a respective new sub-image. S153. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the two-pass classification component (850b), is configured to identify at least two mineral grains in at least one new sub-image based on different intensities (e.g., gray level intensities) between the respective new portions, and wherein the at least two mineral grains are indistinguishable on at least one image (and sub-image), and one of the mineral grains corresponds to a specific mineral grain of a specific portion. S154.2 A system according to any of the preceding embodiments, wherein the path classification component (850b) further includes a second segmentation component. S155.2 A system according to the previous embodiment, wherein the pass classification component (850b), in particular the second segmentation component, is configured to process new sub-images individually when two or more new sub-images are provided. S156.2 A system according to any of the preceding embodiments, wherein the path classification component (850b), in particular the second segmentation component, is configured to generate a new second scan location for each new portion in the at least one new sub-image using a k-means clustering algorithm and a flood-fill algorithm. S157.2 A system according to either of the two preceding embodiments, wherein the path classification component (850b), in particular the second segmentation component, is configured to determine each new second scanning location for the duration of another segmentation period. S158. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the two-pass classification component (850b), is configured to generate third data sets, each of the third data sets including a list of coordinates of new second scan locations associated with one of the new sub-images. S159. A system according to any of the previous embodiments, wherein the data processing system (800), in particular the data storage component (810), is configured to provide at least one of the third data sets. S160. A system according to any of the previous embodiments, wherein the system, in particular the scanning microscope system (100), is configured to focus the beam onto at least one of the new second scan locations of the at least one third data set. S161. A system according to the previous embodiment, wherein the scanning microscope system (100), in particular the second detector (112), is configured to detect a second emission from a new second scan location of a third data set when the sample is irradiated with the beam. S162. A system according to any of the previous embodiments, wherein the scanning microscope system (100), in particular the second detector (112), is configured to detect the second emission for the duration of another third dwell period at each new second scan location of the third data set. S163. A system according to any of the preceding embodiments, wherein the system is configured, if two or more new sub-images are provided, to determine new second scan locations for the new sub-images and detect second emissions from the new second scan locations in the third data set in parallel. S164. A system according to any of the preceding embodiments, wherein the system is configured to determine a new second scan location of the third data set and analyze the first spectrum in parallel. S165. A system according to any of the preceding embodiments, wherein the system is configured to detect a second emission from a second scan location of a third data set and analyze the first spectrum in parallel. S166. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the two-pass classification component (850b), is further configured to generate at least one or more new spectra, each new spectrum based on a second emission detected at a respective new second scan location of the at least one third data set. S167. A system according to the previous embodiment, wherein the data processing system (800), in particular the data storage component (810), is configured to provide new spectra to other groups, each other group including new spectra obtained from at least one or more images. S168. A system according to any of the preceding embodiments, comprising the features of S121-S124, wherein the data processing system (800), in particular the two-pass classification component (850b), is configured to correct image generation errors as performed by the one-pass classification component (850a). S169.2 A system according to any of the preceding embodiments, wherein the path classification component (850b) includes another second spectral analysis component, and wherein the other second spectral analysis component analyzes each other group of new spectra individually. S170. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the other second spectral analysis component, is configured to compare each new spectrum with the reference spectrum and calculate at least two or more other new reliability scores, as performed by the second spectral analysis component. S171. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the two-pass classification component, is configured to perform a second spectral analysis component after the first spectral analysis component has essentially completed analysis of at least some or all of the first spectra associated with at least one or more images. S172. A system according to any of the preceding embodiments, wherein the system is configured to detect a second emission from a new second scan location of the third data set and analyze in parallel a new spectrum associated with the new second scan location. S173. A system according to any of the preceding embodiments, wherein the system is configured to analyze a first other group of new spectra associated with the first image and detect in parallel a second emission from a new second scan location associated with the second image. S174. A system according to any of the preceding embodiments, wherein the system, in particular the two-pass classification component (850b), is configured to identify and classify at least two specific mineral grains within the same new sub-image based on the results of the second segmentation component and the other second spectral analysis component. S175. The system according to either of the previous two embodiments, wherein the third dwell period is longer than the first dwell period. S176. A system according to either of the two preceding embodiments, wherein the other third dwell period is longer than the first dwell period. S177. A system according to any of the previous embodiments, wherein the data processing system (800) further includes a post-processing component (860). S178. A system according to the aforementioned embodiment, wherein the data processing system (800), in particular the post-processing component (860), is configured to acquire other images from other sections of the sample and / or replicant sample at localized contrast and brightness values ​​of at least one of the new sub-images. S179. A system according to any of the preceding embodiments, wherein the data processing system (800), in particular the post-processing component (860), is configured to apply a calibration model, the calibration model being configured to relate chemical compositions of mineral grains (e.g., average atomic number) of the sample to gray level intensities of corresponding portions on each other image. S180. The system according to any of the preceding embodiments, comprising the features of S179, wherein the calibration model corresponds to a linear regression. S181. A system according to any of the preceding embodiments, wherein the calibration model includes contrast and luminance values ​​as model parameters (i.e., fitting parameters). S182. The system according to any of the preceding embodiments, comprising the feature of S179, wherein applying the calibration model includes calibrating and further constraining localized contrast and brightness values ​​using mineral grains of known chemical composition and corresponding gray level intensities in each other image. S183. A system according to any of the preceding embodiments, comprising the features of S182, wherein the data processing system (800), in particular the post-processing component (860), is configured to use the detected gray level intensities and the calibrated contrast and brightness values ​​to determine the chemical composition of unknown mineral grains associated with other sections of the sample and / or replicant sample. S184. A system according to any of the preceding embodiments, wherein the system is a system configured for materials analysis and mineralogy. S185. The system according to any of the previous embodiments, wherein the sample comprises a plurality of particles embedded in an epoxy matrix. S186. A system according to any preceding embodiment, wherein the size of each mineral grain comprises dimensions ranging from at least 1 μm to at most 500 μm. S187. A system according to any of the preceding embodiments, wherein the scan points include dimensions of at most 1 micron. S188. A system according to any preceding embodiment, wherein at least one portion of the image (e.g., a pixel) comprises a dimension in the range of at least 10 nm and at most 1000 nm. S189. A system according to any of the previous embodiments, wherein the second dwell period is in the range of at least 1 millisecond to at most 10 milliseconds. S190. The data processing system (800), in particular the one-pass classification component (850a), averages at least 2 x 10 per second scan location. 3 photons, up to 3x10 3 3. A system according to any of the previous embodiments, configured to calculate a high confidence score for at least some or all mineral particles in a sample by detecting photons. S191. A system according to any of the preceding embodiments, wherein the system further includes a control unit (115), the control unit (115) configured to control the power supply and operation of several components of the scanning microscope system (100), such as the focusing lens (104), the objective lens (106), the scanning coil (105), and the movable stage (113). S192. A system according to any of the previous embodiments, wherein the system further comprises a vacuum system, the vacuum system comprising a vacuum controller (116), a mechanical pumping system (117), an ultra-high vacuum pump (118), and a vacuum chamber (119). S193. A system according to any of the previous embodiments, wherein the mechanical pumping system (117) and the ultra-high vacuum pump (118) are configured to provide an ultra-high vacuum in the vacuum chamber (119). S194. A system according to the previous embodiment, wherein the vacuum chamber (119) is configured to contain the sample (108), the movable stage (113), the first detector (111), the second detector (112) or portions thereof, and the scanning electron microscope (101) or portions thereof.

[0412] Method Embodiments Discussed below are method embodiments, which are abbreviated by the letter "M" followed by a number. These embodiments are meant whenever reference is made herein to "method embodiments." M1. A method for determining the characteristics of a sample or a section thereof - providing at least one or more images of the sample or sections thereof based on first emissions detected within a first dwell period from a plurality of first scanning locations; - performing a first detection step, the first detection step including detecting second emissions during a second dwell period from at least one or more second scan locations of at least one region of at least one image, each second scan location associated with a portion of a corresponding region; - performing a first spectrum providing step, the first spectrum providing step including providing at least one or more first spectra, each first spectrum being based on second emissions detected at each of second scan locations of the at least one region; - performing a first spectral analysis step, the first spectral analysis step including calculating a reliability score for all first spectra and selecting second scan locations associated with first spectra having a reliability score below a threshold; - performing a classification step, the classification step including detecting a second emission from at least one of the selected second scan locations during a third dwell period and / or providing at least one or more new images delimiting portions associated with the selected second scan locations, and determining a new second scan location in the corresponding new image via its modified contrast and brightness values ​​for the at least one image. M2. A method according to any preceding embodiment, wherein the method further comprises a first segmentation step, the first segmentation step comprising determining a second scanning location of the at least one region of the image. M3. A method according to any of the previous embodiments, further comprising focusing a beam of charged particles (such as electrons) to a scan point on the sample. M4. A method according to any of the preceding embodiments, wherein the method further comprises scanning the beam of charged particles across a plurality of scan locations in one or two dimensions. M5. The method according to any of the preceding embodiments, wherein the scan location corresponds to the first scan location. M6. The method according to any of the preceding embodiments, wherein the scan location corresponds to the second scan location. M7. A method according to any of the preceding embodiments, wherein the method further comprises assigning a two-dimensional coordinate system to the sample. M8. The method according to any of the preceding embodiments, comprising the features of M7, wherein the method further comprises assigning a two-dimensional coordinate system of the sample to at least one image. M9. A method according to any of the preceding embodiments, comprising the features of M4, wherein the result of scanning the beam of charged particles across the scanned locations of the sample includes an interaction of the beam with the sample. M10. A method according to any preceding embodiment, comprising the features of M1, wherein the result of the interaction includes a first and / or second emission. M11. A method according to any preceding embodiment, comprising features of M1 and M10, wherein the first emission comprises an emission of particles (such as backscattered electrons). M12. A method according to any of the preceding embodiments, comprising features of M1 and M10, wherein the second emission comprises an emission of photons (such as X-ray photons). M13. A method according to any of the preceding embodiments, wherein the method further comprises detecting a first emission from each first scan location. M14. A method according to any preceding embodiment, wherein the method further includes generating at least one image based on the first emissions detected at each first scan location. M15. A method according to any of the previous embodiments, wherein at least one image corresponds to a backscattered electron image. M16. A method according to any of the preceding embodiments, wherein at least one image shows intensity variations between regions (and / or portions thereof) having different properties (such as chemical composition). M17. A method according to any preceding embodiment, wherein the intensity variation comprises a gray level variation. M18. A method according to any of the preceding embodiments, wherein at least one image comprises contrast and brightness values. M19. A method according to any of the preceding embodiments, wherein each region of at least one image corresponds to a particle in the sample. M20. A method according to any preceding embodiment, wherein each particle in the sample comprises at least one or more mineral grains. M21. A method according to any of the preceding embodiments, wherein the method further comprises detecting a first emission at each first scan location for the duration of a first dwell period. M22. A method according to any of the preceding embodiments, wherein the first detecting step further comprises detecting a second emission at each second scan location for the duration of a second dwell period. M23. A method according to any of the preceding embodiments, comprising the features of M1, wherein the second dwell period is longer than the first dwell period. M24. The method according to any of the preceding embodiments, further comprising detecting the first emission and detecting the second emission at different time intervals, wherein the different time intervals correspond to non-overlapping time intervals. M25. A method according to any of the preceding embodiments, comprising the features of M2, wherein the first segmentation step includes determining each second scan location for the duration of the segmentation period. M26. A method according to any of the preceding embodiments, wherein the segmentation period is longer than the second dwell period. M27. A method according to any of the preceding embodiments, wherein the segmentation period is less than or equal to the second dwell period. M28. A method according to any of the preceding embodiments, wherein the segmentation period depends on image characteristics such as resolution and magnification of at least one image (or section thereof). M29. A method according to any of the previous embodiments, wherein the segmentation period depends on the size of the mineral grains and / or particles. M30. A method according to any of the preceding method embodiments, further comprising a pre-processing step. M31. A method according to any preceding embodiment, wherein the pre-processing step comprises applying a thresholding algorithm. M32. A method according to any preceding embodiment, wherein the thresholding algorithm includes separating at least one image into background and foreground portions based on a threshold strength. M33. A method according to any of the preceding embodiments, wherein the background portion comprises a background portion, and the background portion (e.g., pixels) comprises an intensity lower than a threshold intensity (e.g., dark gray and / or black portion). M34. The method according to any of the preceding embodiments, wherein the foreground portion comprises a foreground portion, and the foreground portion (e.g., pixel) comprises an intensity equal to or greater than a threshold intensity (e.g., light gray and / or white portion). M35. A method according to any of the previous embodiments, wherein the thresholding algorithm includes segmenting / removing background portions. M36. A method according to any preceding embodiment, wherein segmenting / removing background portions comprises assigning the background portions a color and / or intensity of the same value (e.g., black portions). M37. A method according to any of the preceding embodiments, wherein the foreground portion comprises at least a portion of an area of ​​the image. M38. A method according to any preceding embodiment, wherein the preprocessing step comprises determining the boundaries of the corresponding regions of the at least one image using a contour algorithm. M39. A method according to any preceding embodiment, wherein the contour algorithm includes joining adjacent portions along the boundaries of the corresponding regions into curves. M40. A method according to the preceding embodiment, wherein adjacent portions along the boundary of the corresponding region are surrounded by background portions (e.g., black portions). M41. A method according to any of the preceding embodiments, wherein the preprocessing step further comprises applying a bounding box algorithm, the bounding box algorithm comprising dividing the at least one image into at least one or more sub-images based on the results of the contour algorithm. M42. A method according to any of the preceding embodiments, comprising the features of M41, wherein a sub-image of at least one image delimits a region. M43. A method according to any of the preceding embodiments, comprising the features of M41, wherein the method includes providing a sub-image. M44. A method according to any of the preceding embodiments, wherein the first segmentation step includes correcting sub-image generation errors. M45. A method according to any preceding embodiment, wherein the sub-image generation error includes generating at least one sub-image that includes at least two adjacent regions. M46. A method according to any of the preceding embodiments, wherein at least two adjacent regions located within one sub-image correspond to touching particles within the sample. M47. A method according to any of the preceding embodiments, wherein at least one or more portions along a boundary of one of the adjacent regions is contiguous with at least one or more portions along a boundary of another of the adjacent regions. M48. A method according to any of the preceding embodiments, having the features of M44, wherein correcting sub-image generation errors includes processing each of adjacent regions within one sub-image separately. M49. A method according to any of the preceding embodiments, having the features of M41, wherein if two or more sub-images are provided, the method further comprises performing the first segmentation step on the sub-images individually. M50. A method according to any of the preceding embodiments, wherein the first segmentation step includes assigning contiguous portions of parts of corresponding regions in each sub-image to clusters using a k-means clustering algorithm. M51. A method according to either of the previous two embodiments, wherein each region of each sub-image comprises at least one cluster. M52. A method according to any of the preceding embodiments, having the features of M50, wherein the first segmentation step further comprises applying a flood-fill algorithm, the flood-fill algorithm comprising generating a mask for at least one of the clusters. M53. A method according to any preceding embodiment, wherein generating a mask for at least one of the clusters includes assigning the same value of color and / or intensity to consecutive portions within corresponding clusters. M54. A method according to any of the preceding embodiments, wherein each mask delimits a portion of a corresponding region. M55. A method according to any preceding embodiment, wherein portions within each region of the at least one image correspond to mineral grains within a corresponding grain in the sample. M56. A method according to any of the preceding embodiments, wherein the first segmentation step further comprises determining one second scan location for each mask. M57. A method according to any of the preceding embodiments, wherein each second scan location corresponds to a centroid of a respective mask. M58. A method according to any of the preceding embodiments, wherein the method further comprises correcting over-segmentation errors. M59. A method according to any preceding embodiment, wherein the over-segmentation error includes determining two or more second scan locations for at least one of the masks. M60. A method according to either of the two previous embodiments, comprising merging second scan locations for at least one of the masks into one second scan location by a merge operator. M61. A method according to any of the preceding embodiments, wherein the method further comprises generating a first data set for a region within each sub-image. M62. A method according to any preceding embodiment, wherein the first data set includes a list of coordinates of second scan locations associated with one of the regions. M63. A method according to any of the preceding embodiments, wherein the method further comprises providing at least one of the first data sets. M64. A method according to any of the preceding embodiments, wherein the first detecting step includes focusing a beam onto at least one of the second scan locations of at least one of the first data sets. M65. A method according to any preceding embodiment, wherein the first detecting step further comprises detecting a second emission from a corresponding second scan location of the first data set when the sample is irradiated with the beam. M66. A method according to any of the preceding embodiments, wherein the method further comprises performing the first segmentation step and the first detection step in parallel if two or more sub-images (i.e., first data sets) are provided. M67. A method according to any of the preceding embodiments, comprising features of M1, wherein the method includes generating a first spectrum based on a second emission (i.e., number of photons) detected at a corresponding second scan location of at least one sub-image (i.e., first data set). M68. A method according to any of the preceding embodiments, wherein each first spectrum corresponds to an X-ray spectrum. M69. A method according to the preceding embodiment, wherein the X-ray spectrum comprises at least one or more spectral lines. M70. The method according to any of the preceding embodiments, comprising features of M68 and M69, wherein the X-ray spectrum comprises a number of detected X-ray photons at each energy (i.e., spectral line intensities). M71. A method according to any preceding embodiment, wherein each spectral line corresponds to an electronic transition of a chemical element. M72. A method according to any of the preceding embodiments, wherein each mineral grain of the sample comprises at least one or more chemical elements. M73. The method according to any of the preceding embodiments, wherein the X-ray spectrum includes information regarding the chemical composition (e.g., mineral composition) of corresponding mineral grains associated with each second scan location. M74. A method according to any of the preceding embodiments, wherein the first spectral analysis step includes analyzing each first spectrum from a respective second scan location of at least one of the regions (i.e., the first data set). M75. A method according to any preceding embodiment, wherein analyzing each first spectrum includes comparing each first spectrum to at least one or more reference spectra. M76. A method according to any preceding embodiment, wherein each reference spectrum includes a plurality of predetermined spectral lines associated with known mineral grains. M77. A method according to any of the preceding embodiments, wherein the first spectral analysis step further includes a first line assignment step, the first line assignment step including assigning spectral lines of each first spectrum to predetermined spectral lines of the reference spectrum. M78. A method according to any of the preceding embodiments, wherein the first spectral analysis step includes matching each first spectrum to one of the known mineral grains based on the results of the first line assignment step. M79. A method according to any preceding embodiment, comprising the features of M1, wherein the confidence score corresponds to a level of match between the first spectrum and the corresponding matched reference spectrum. M80. The method according to any of the previous embodiments, wherein a reliability score corresponds to a numerical value, the numerical value ranging from 0 to at most 1, assigned to each first spectrum. M81. The method according to any of the preceding embodiments, wherein the confidence score represents the probability that each first spectrum belongs to one of the known mineral grains. M82. The method according to any of the preceding embodiments, having the features of M1, wherein the method includes presetting a threshold for the reliability score. M83. The method according to any of the preceding embodiments, comprising the features of M82, wherein a high confidence score (above a threshold) corresponds to a reliable identification of the chemical composition of each mineral grain based on the corresponding first spectrum. M84. The method according to any of the preceding embodiments, comprising the features of M82, wherein a low confidence score (below a threshold) corresponds to partial identification of the chemical composition of each mineral grain based on the corresponding first spectrum. M85. A method according to any of the preceding embodiments, wherein a mineral grain with a low confidence score comprises a similar chemical composition to at least another mineral grain. M86. A method according to any preceding embodiment, wherein mineral grains of similar chemical composition contain at least one or more common chemical elements. M87. A method according to any of the preceding embodiments, wherein mineral grains of similar chemical composition contain the same chemical elements in different elemental ratios (e.g., Fe2O3 and Fe3O4). M88. The method according to any of the preceding embodiments, characterized by M86, wherein mineral grains of similar chemical composition correspond to similar first spectra, and the similar first spectra include at least one or more common spectral lines. M89. The method according to any of the preceding embodiments, characterized in that the similar first spectra include the same spectral lines (at the same energy) having different intensity ratios. M90. The method according to any of the preceding embodiments, comprising the features of M84, wherein the low confidence score is due to low spectral quality of the respective first spectrum. M91. A method according to any preceding embodiment, wherein the poor spectral quality results from spectral lines of each first spectrum having insufficient spectral line intensity, and the insufficient line intensity results from an insufficient number of photons detected during the second dwell period. M92. A method according to any of the preceding embodiments, comprising estimating an intrinsic photon count rate based on the number of photons detected during the second dwell period (i.e., the number of detected photons per time period). M93. A method according to any of the preceding embodiments, characterized in that the characteristic photon count rate depends on the chemical composition of the corresponding moiety (i.e., mineral grain). M94. The method according to any of the preceding embodiments, characterized in M93, wherein the characteristic photon count rate depends on sample properties such as crystalline properties (e.g., orientation, size, depth) of the corresponding portion (i.e., mineral grain). M95. The method according to any of the preceding embodiments, wherein the method further comprises generating and providing at least one or more second data sets, each second data set comprising a list of coordinates of second scan locations associated with mineral grains with low confidence scores in at least one or more regions (i.e., sub-images) of the at least one image. M96. A method according to any of the preceding embodiments, wherein the method includes performing the first detecting step and the first spectrum providing step in parallel. M97. A method according to any of the preceding embodiments, wherein the method comprises performing the first detection step and the first spectral analysis step in parallel. M98. A method according to any of the preceding embodiments, wherein the classification step comprises a one-pass classification step and / or a two-pass classification step. M99. A method according to any of the preceding embodiments, wherein the classification step includes performing a one-pass and / or two-pass classification step based on the results of the first spectral analysis step. A method according to any of the preceding embodiments, wherein the M100.1-pass classification step includes performing a one-pass and / or two-pass classification step after a first detection step has essentially been performed on the corresponding image. A method according to any of the preceding embodiments, wherein the M101.1 pass classification step includes performing a one-pass and / or two-pass classification step after the first detection step is completed. M102.1 A method according to any of the preceding embodiments, wherein the path classification step includes a second detection step, and the second detection step includes focusing the beam on at least one of the second scan locations of the second data set. M103. A method according to any preceding embodiment, wherein the second detecting step further comprises detecting a second emission from each second scan location of the second data set when the sample is irradiated with the beam. M104. A method according to any of the preceding embodiments, wherein the second detecting step comprises detecting a second emission from at least one second scan location (of the second data set) for the duration of a third dwell period. M105. A method according to any of the preceding embodiments, wherein for at least one second scan location, the third dwell period is longer than the second dwell period. M106. A method according to any of the preceding embodiments, wherein for at least one second scan location, the third dwell period is less than or equal to the second dwell period. M107.1 The method according to any of the preceding embodiments, comprising the feature of M92, wherein the path classification step includes determining a third dwell period for at least one second scanning location of the second data set based on the calculated confidence score of each first spectrum and the characteristic photon count rate of the corresponding mineral grain. M108.1 A method according to any of the preceding embodiments, wherein the path classification step includes adding the number of photons (e.g., X-ray photons) detected by the first detection step to the number of photons (e.g., X-ray photons) detected by the second detection step at each second scanning location of the second data set. M109.1 A method according to any of the preceding embodiments, comprising the features of M108, wherein the path classification step further comprises generating at least one or more second spectra, each second spectrum being based on second emissions detected by the first and second detection steps at a respective second scan location of the second data set. M110. A method according to any of the preceding embodiments, comprising the features of M109, wherein the method includes a second spectrum providing step, the second spectrum providing step including providing the second spectra in groups, each group including second spectra obtained from at least one or more images. M111. A method according to any of the preceding embodiments, wherein each second spectrum corresponds to an X-ray spectrum. M112.1 A method according to any of the preceding embodiments, comprising the features of M110, wherein the path classification step includes correcting image generation errors if the group includes a second spectrum obtained from at least two images, the at least two images showing adjacent sections of the sample. M113. A method according to any preceding embodiment, wherein at least two images include at least two portions belonging to one of the mineral grains with a low confidence score. M114. A method according to any of the preceding embodiments, comprising the features of M112, wherein correcting image generation errors comprises stitching at least two portions of at least two images. M115.1 The method according to any of the preceding embodiments, having the feature of M114, wherein the path classification step includes summing, using another merge operator, second spectra of parts belonging to the same mineral grain with a lower confidence score. M116.1 A method according to any of the preceding embodiments, wherein the path classification step includes a second spectral analysis step. M117.1 A method according to any of the preceding embodiments, wherein the path classification step includes performing a second spectral analysis step separately on the second spectra of each group. M118. A method according to any of the preceding embodiments, wherein the second spectral analysis step includes matching each second spectrum to at least two known mineral grains based on the results of the second line assignment step. M119. A method according to any of the preceding embodiments, wherein the second spectral analysis step includes calculating at least two or more new confidence scores for each second spectrum, each new confidence score corresponding to a level of match between the second spectrum and a corresponding matched reference spectrum. M120.1 A method according to any of the preceding embodiments, having the features of M119, wherein the path classification step includes selecting a highest new confidence score from among the at least two new confidence scores for each second spectrum. M121. A method according to any of the previous embodiments, wherein the highest new confidence scores of at least some of the second spectra correspond to high confidence scores (above a threshold). A method according to any of the preceding embodiments, comprising the features of M121, wherein the path classification step includes normalizing data quality for mineral grains with (initial) low confidence scores. M123.1 A method according to any of the preceding embodiments, wherein the path classification step includes performing the second detection step and the first spectral analysis step in parallel. A method according to the preceding embodiment, wherein the path classification step includes performing a second spectral analysis step essentially after a first spectral analysis step has been performed on at least one or more images. M125.1 A method according to any of the preceding embodiments, comprising the features of M123 and M124, wherein the path classification step includes performing the second detection step and the second spectral analysis step in parallel. A method according to the preceding embodiment, wherein the M126.1 path classification step includes performing a second spectral analysis step on a first group of second spectra associated with the first image while concurrently performing a second detection step on a second scan location associated with the second image. M127. The method according to any of the preceding embodiments, wherein at least one or more of the mineral grains with low confidence scores associated with the second dataset corresponds to a specific mineral grain. M128. A method according to any of the preceding embodiments, wherein at least one of the specific mineral grains is depicted at the same or similar intensity (i.e., gray level intensity) on at least one image as at least one other specific mineral grain (e.g., Fe2O3 and Fe3O4). M129. A method according to any of the preceding embodiments, wherein at least one particular mineral grain is indistinguishable from at least one other particular mineral grain on at least one image. M130. A method according to any of the preceding embodiments, wherein the portion of the selected second scan location associated with a particular mineral grain corresponds to a particular portion of the respective sub-image. M131. A method according to any of the preceding embodiments, wherein the classification step further comprises performing a two-pass classification step on selected second scanned locations of the second data set associated with particular portions of the corresponding sub-images. A method according to any of the preceding embodiments, wherein the M132.2 path classification step includes presetting and / or controlling contrast and brightness values ​​of the corresponding image (e.g., at least one image or section thereof) by adjusting operational settings of at least one or more system components prior to detecting the corresponding first emission. M133.2 A method according to any of the preceding embodiments, wherein the path classification step further comprises focusing the beam to a first scan location within the particular portion. M134.2 A method according to any of the preceding embodiments, comprising the features of M131, wherein the path classification step includes redetecting a first emission from a first scan location of a corresponding specific portion once the beam has illuminated the sample. M135. A method according to any of the preceding embodiments, wherein the method comprises generating and providing at least one or more new sub-images, each new sub-image being based on the first emission redetected in each particular portion. A method according to any of the preceding embodiments, comprising the features of M1 and M132, wherein the M136.2 path classification step further comprises obtaining a new sub-image having adjusted contrast and brightness values ​​by correspondingly pre-adjusting an operational setting of at least one system component. M137. A method according to any of the preceding embodiments, comprising the features of M1, wherein the new image corresponds to a new sub-image. M138. A method according to any of the preceding embodiments, comprising the features of M1, wherein the portion bounded by the new image corresponds to a particular portion bounded by the new sub-image. A method according to any of the preceding embodiments, wherein the M139.2 path classification step further comprises revealing and / or detecting at least two or more new portions in at least one of the new sub-images by means of the adjusted contrast and brightness values ​​of each new sub-image. M140. A method according to any of the preceding embodiments, comprising the features of M139, wherein each new portion corresponds to a section of a particular portion within a respective new sub-image. M141.2 A method according to any of the preceding embodiments, comprising the features of M139, wherein the path classification step includes identifying at least two mineral grains in at least one new sub-image based on different intensities (e.g., gray level intensities) between each new portion, and wherein the at least two mineral grains are indistinguishable on at least one image (and sub-image), and one of the mineral grains corresponds to a specific mineral grain of a specific portion. A method according to any of the preceding embodiments, wherein the M142.2 path classification step further comprises, if two or more new sub-images are provided, performing a second segmentation step on the new sub-images individually. M143. A method according to any of the preceding embodiments, wherein the second segmentation step includes generating new second scan locations for each new portion in the at least one new sub-image using a k-means clustering algorithm and a flood-fill algorithm. M144. A method according to either of the two preceding embodiments, wherein the second segmentation step includes determining each new second scanning location for the duration of another segmentation period. M145.2 A method according to any of the preceding embodiments, wherein the path classification step includes generating third data sets, each of the third data sets including a list of coordinates of new second scan locations associated with one of the new sub-images. M146. The method according to any of the preceding embodiments, wherein the method further comprises providing at least one of the third data sets. M147.2 A method according to any of the preceding embodiments, wherein the path classification step includes performing another second detection step on a new second scanning location of the at least one third data set. M148. A method according to any of the preceding embodiments, wherein the further second detecting step includes focusing the beam on at least one of the new second scan locations of the at least one third data set. M149. A method according to any preceding embodiment, wherein the other second detection step further comprises detecting a second emission from a corresponding new second scan location when the sample is irradiated with the beam. M150. The method according to any of the preceding embodiments, wherein the other second detection step includes detecting a second emission for the duration of another third dwell period for each new second scan location of each third data set. A method according to any of the preceding embodiments, wherein the M151.2 path classification step includes performing a second segmentation step and another second detection step in parallel if two or more new sub-images are provided. A method according to any of the preceding embodiments, wherein the M152.2 path classification step includes performing the second segmentation step and the first spectral analysis in parallel. M153.2 A method according to any of the preceding embodiments, wherein the path classification step includes performing another second detection step and a first spectral analysis in parallel. M154.2 A method according to any of the preceding embodiments, wherein the path classification step further comprises generating at least one or more new spectra, each new spectrum based on a second emission detected by the other second detection step at a respective new second scanning location of the at least one third data set. M155. A method according to any preceding embodiment, wherein the method includes a new spectrum providing step, wherein the new spectrum providing step includes providing new spectra in other groups, each other group including new spectra obtained from at least one or more images. M156. A method according to any of the preceding embodiments, including features M112-M115, wherein the two-pass classification step includes correcting image generation errors as performed in a one-pass classification step. A method according to any of the preceding embodiments, wherein the M157.2 path classification step includes performing a separate second spectral analysis step on each other group of new spectra individually. M158. A method according to any of the preceding embodiments, having the features of M119-M121, including comparing each new spectrum to a reference spectrum and calculating at least two or more other new confidence scores, as performed in the second spectral analysis step of the one-pass classification step. A method according to any of the preceding embodiments, wherein the M159.2 path classification step includes performing a further second spectral analysis step after the first spectral analysis has essentially been performed on at least one or more images. A method according to any of the preceding embodiments, wherein the M160.2 path classification step includes performing another second detection step and another second spectral analysis step in parallel. A method according to the preceding embodiment, wherein the M161.2 path classification step includes performing another second spectral analysis on a first other group of new spectra associated with the first image and performing a second detection step in parallel on a new second scanning location associated with the second image. A method according to any of the preceding embodiments, wherein the M162.2 pass classification step includes identifying and classifying at least two specific mineral grains within the same new sub-image based on the results of the second segmentation step and the other second spectral analysis. M163. The method according to either of the previous two embodiments, wherein the third dwell period is longer than the first dwell period. M164. The method according to either of the previous two embodiments, wherein the other third dwell period is longer than the first dwell period. M165. A method according to any of the preceding embodiments, wherein the method further comprises a post-processing step. M166. A method according to the preceding embodiment, wherein the post-processing step comprises obtaining other images from other sections of the sample and / or replicant sample at localized contrast and brightness values ​​of at least one of the new sub-images. M167. A method according to any of the preceding embodiments, wherein the post-processing step includes applying a calibration model, the calibration model including relating the chemical composition of the mineral grains of the sample (e.g., average atomic number) to the gray level intensity of corresponding portions on each other image. M168. The method according to any of the preceding embodiments, having the features of M167, wherein the calibration model corresponds to a linear regression. M169. A method according to any of the preceding embodiments, wherein the calibration model includes contrast and brightness values ​​as model parameters (i.e., fitting parameters). M170. The method according to any of the preceding embodiments, having the features of M167, wherein applying the calibration model includes calibrating and further constraining localized contrast and brightness values ​​using mineral grains of known chemical composition and corresponding gray level intensities in each other image. M171. A method according to any of the preceding embodiments, comprising the features of M170, wherein a post-processing step includes using the detected gray level intensities and the calibrated contrast and brightness values ​​to determine the chemical composition of mineral grains associated with other sections of the sample and / or replicant sample. M172. A method according to any of the preceding method embodiments, wherein at least a portion of the one-pass classification step and at least a portion of the two-pass classification step are computer-implemented. M173. A method according to any of the preceding embodiments, wherein the pre-processing step, the first segmentation step, the first spectral analysis step, and the post-processing step are computer-implemented. M174. A method according to any of the preceding embodiments, wherein the method is a method for materials analysis and mineralogy. M175. A method according to any of the preceding embodiments, wherein the sample comprises a plurality of particles embedded in an epoxy matrix. M176. A method according to the previous embodiment, wherein the size of each mineral grain comprises a dimension ranging from at least 1 μm to at most 500 μm. M177. A method according to any of the preceding embodiments, wherein the scan points comprise dimensions of at most 1 micron. M178. A method according to the preceding embodiment, wherein at least one portion of the image (such as a pixel) comprises a dimension in the range of at least 10 nm and at most 1000 nm. M179. A method according to any of the preceding embodiments, wherein the second dwell period is in the range of at least 1 millisecond to at most 10 milliseconds. M180.1 pass classification steps average at least 2 x 10 per second scan location 3 photons, up to 3×10 3 3. The method according to any of the preceding embodiments, comprising calculating a high confidence score for at least some or all of the mineral grains in the sample by detecting photons. M181. A method according to any of the preceding method embodiments, comprising using a system according to any of the system embodiments. S195. A system according to any of the preceding system embodiments, wherein the system is configured to perform a method according to any of the preceding method embodiments.

[0413] In the following, computer program product embodiments are discussed. These embodiments are abbreviated by the letter "C" followed by a number. These embodiments are meant whenever reference is made herein to "computer program product embodiments." C1. A computer program product comprising instructions which, when the program is executed by a computer, cause the computer to perform the steps of the method according to any of the method embodiments. C2. A computer program product comprising instructions that, when executed by a data processing system (800), cause the data processing system (800) to perform steps, wherein the data processing system (800) of the system according to any of the system embodiments is configured for the steps.

[0414] Exemplary features of the present invention are further detailed in the figures and the following figure descriptions. [Brief explanation of the drawings]

[0415] [Figure 1] 1 shows a scanning microscope system. [Figure 2] A sample of particles is shown. [Figure 3] An image of the sample and a sub-image of the polymineral particles are shown. [Figure 4] Shown are a sub-image of a polymineral grain and a new sub-image of a mineral grain with modified contrast and brightness settings. [Figure 5] 1 illustrates an embodiment of a method. [Figure 6] 1 shows a first flowchart depicting another embodiment of the method. [Figure 7a] 7 shows a second flowchart depicting the steps of the one-pass classification process of the embodiment of FIG. 6. [Figure 7b] 7 shows a third flowchart depicting the steps of the two-pass classification process of the embodiment of FIG. 6. [Figure 8] 1 shows a system configured to perform the method. DETAILED DESCRIPTION OF THE INVENTION

[0416] For clarity, some features may be shown in only some figures and other features may be omitted, however, omitted features may be present and the features depicted and discussed need not be present in all embodiments.

[0417] FIG. 1 illustrates components of a scanning microscope system 100. The scanning microscope system 100 may be configured to generate a primary beam of charged particles (e.g., electrons or ions). The scanning microscope system may further include a scanning electron microscope 101. In this example, the primary beam includes an electron beam 107. The electron source 102 may be configured to emit an electron beam, where a voltage is applied between the electron source 102 and an anode 103. The applied voltage may preferably be in the range of at least 2 kV to a maximum of 30 kV. The scanning microscope system may also include an electromagnetic lens. The electromagnetic lens may be configured to control the path of the electron beam. The at least one condenser lens 104 may be configured by an electromagnetic lens. The condenser lens 104 may be configured to determine the size of the electron beam. Furthermore, the at least one objective lens 106 may be configured by an electromagnetic lens. The objective lens 106 may be configured to focus the electron beam to a scanning point on a sample. The scanning point may correspond to an electron spot on the sample 108. Furthermore, the size and shape of the scanning point may depend on the focusing characteristics of the electromagnetic lens (e.g., applied current) and the working distance between the scanning electron microscope 101 and the sample 108. The scanning coil 105 may be configured to deflect the electron beam 107 over multiple scanning locations in one or two dimensions. Thus, optionally and advantageously, two-dimensional scanning of the sample may be possible. The scanning location may correspond to a first scanning location. The scanning location may also correspond to a second scanning location. The scanning coil 105 may be magnetic or electrostatic.

[0418] The scanning microscope system can be configured to generate first and second emissions 109, 110. The electron beam 107 can interact with particles (such as atoms) of the sample 108. The interaction can result in the first and second emissions 109, 110. The first emission 109 can include an emission of charged particles, such as backscattered electrons. However, the first emission can also include an emission of secondary electrons, transmitted electrons, and / or Auger electrons. Additionally, the second emission 110 can include an emission of photons, such as X-rays and / or light (e.g., visible light).

[0419] The scanning microscope system 100 may also include a first detector 111, which may be configured to continuously detect the first emissions 109 from the first scan locations. In particular, the first detector 111 may be configured to detect the first emissions for a first dwell period at each first scan location.

[0420] The first detector 111 may include a backscattered electron detector, such as a segmented silicon drift detector. However, the backscattered electron detector may also correspond to other types of solid-state detectors. Furthermore, the first detector 111 may also include a secondary electron detector, such as an Everhart-Thornley detector, or a transmission electron detector (e.g., a CMOS detector). The transmission electron detector may be positioned below the sample 108 to detect transmitted electrons.

[0421] Additionally, the electron microscope system may include a second detector 112 configured to continuously detect the second emission 110 from the second scanning location. In particular, the second detector 112 may be configured to detect the second emission 110 over a second dwell period at each second scanning location. The second detector 112 may include an X-ray detector, which may include a silicon drift detector. However, the X-ray detector may also include other types of detectors (e.g., scintillation detectors). The second detector 112 may be tilted relative to the surface of the sample 108. The angle between the centerline 114 of the second detector and the sample surface may be adjustable and may range from 0° to a maximum of 90°.

[0422] The X-ray detector may be an energy dispersive spectrometer (EDS). The energy bandwidth of the EDS may range from 0 to a maximum of 17 keV. In another modality, the X-ray detector may be a wavelength dispersive spectrometer (WDS). Furthermore, the second detector 112 may also be an electron energy loss spectrometer or a cathodoluminescence spectrometer.

[0423] The sample 108 may be positioned on top of a movable stage 113. The movable stage 113 may be configured to perform two horizontal movements, a vertical movement, a tilt movement, and / or a rotation movement relative to the plane of the sample. The two horizontal movements may include selecting a field of view. The vertical movement may include changing the height of the sample and therefore the depth of focus and / or image resolution.

[0424] The scanning microscope system 100 may further include a control unit 115. The control unit 115 may be configured to control the power supply and operation of the focusing lens 104, the objective lens 106, the scanning coil 105, and the movable stage 113. Furthermore, the scanning microscope system may include a vacuum system. The vacuum system may include a vacuum controller 116, a mechanical pumping system 117, an ultra-high vacuum pump 118 (such as an ion pump), and a vacuum chamber 119. The vacuum controller 116 may be configured to control the operation of the mechanical pumping system 117 and the ultra-high vacuum pump 118. The mechanical pumping system 117 and the ultra-high vacuum pump 118 may be configured to provide an ultra-high vacuum within the vacuum chamber 119. The vacuum chamber 119 may be configured to house the sample 108, the movable stage 113, the first detector 111, the second detector 112, or portions thereof, and the scanning electron microscope 101, or portions thereof.

[0425] FIG. 2 illustrates a sample 108 including multiple particles 200, each of which may include at least one mineral grain. The particles 200 may be embedded in an epoxy matrix 201. An internal two-dimensional coordinate system (x, y) may be assigned to the sample 108. In this example, the origin of the internal coordinate system may be assigned to the lower left corner of the sample. Additionally, FIG. 2 illustrates a first scan location 202. The sequence of first scan locations may correspond to a scan pattern. As seen in FIG. 2, the scan pattern may correspond to a quadratic meandering pattern. The quadratic meandering pattern may include a large meandering pattern moving back and forth along a large row (dashed line 203) and a small meandering pattern moving up and down within every single row. However, the scan pattern may also correspond to a continuous fractal space or plane-filling curve, such as a Hilbert curve or a Moore curve.

[0426] FIG. 3 shows an image 300 of the sample 108. The image may include regions 301, where each region may correspond to a particle 200 of the sample. The image 300 may be generated based on the first emission 109 detected at a corresponding first scan location 202. In this example, the image may be generated based on the emission of backscattered electrons. The image may include a two-dimensional grid, with each square of the grid corresponding to a portion (e.g., a pixel). Each portion of the image may correspond to a scan point on the sample. Backscattered electron data may be stored point-by-point in each portion of the image. Thus, the intensity of each portion may depend on the number of backscattered electrons detected at each scan point. Furthermore, the same internal two-dimensional coordinate system (x, y) of the sample 108 may also be assigned to the image 300. Thus, the location of each portion within the image may be tracked as the electron beam moves across the first scan location of the sample. Assigning the same coordinate system of the sample to the image may be achieved using a reference point of known coordinates, which may be incorporated into the sample 108 or the movable stage 113.

[0427] 3 further illustrates a sub-image 302 including a region 301, where the sub-image 302 is a section of the image 300. The region 301 of the sub-image 302 may include portions 303, each of which may correspond to a mineral grain within the corresponding particle 200. Furthermore, each portion may include portions of the same color (e.g., gray) and / or intensity. In other words, the sub-image may show intensity variations between portions (e.g., mineral grains) with different chemical compositions. The intensity variations may include gray level variations. Furthermore, the image 300 may include contrast and brightness values.

[0428] 3, each portion 303 includes a second scan location 304, which may correspond to the centroid or another estimate of the corresponding portion. The first and second scan locations may be specified with respect to the same two-dimensional coordinate system (x, y).

[0429] FIG. 3 further depicts a spectrum 305, which may be generated based on the second emission 110 detected at the corresponding second scan location 304. In this example, the spectrum may include a graph depicting the number of detected X-ray photons (i.e., intensity) at each energy. As seen in FIG. 3, the X-ray spectrum may include spectral lines 306 at different energy values. Furthermore, the spectral lines 306 may include a linewidth and thus may be distributed across a range of energy values ​​rather than being located at a single energy value. As such, different spectral lines associated with the same and / or different elements may overlap. The peak of each spectral line 306 at the center of the corresponding line may correspond to the maximum number of detected X-ray photons (i.e., peak intensity) across the corresponding linewidth. Each spectral line may correspond to an electronic transition of a chemical element (e.g., silicon, iron, etc.), and the energy value of each electronic transition may be unique to the corresponding chemical element. Furthermore, the X-ray spectrum may include spectral lines associated with at least one or more chemical elements (e.g., mineral elements). Thus, optionally and advantageously, the spectrum may include information regarding the chemical composition (eg, mineral composition) of the corresponding second scan location associated with the mineral grain.

[0430] FIG. 4 shows a new sub-image 400 that delimits one of the portions 303 of the sub-image 302. The new sub-image thus comprises a section of the sub-image 302. The delimited portion 303 may correspond to a specific portion, which may relate to a specific mineral grain (e.g., Fe2O3). The new sub-image includes contrast and brightness values ​​adjusted relative to the sub-image 302, thereby revealing two new portions 401. In other words, the two new portions 401 may be sections of the (specific) portion 303 that include different gray level intensities from each other and from the portion 303. The two new portions 401 may correspond to two mineral grains having similar chemical compositions (e.g., Fe2O3 and Fe3O4). One of the mineral grains may correspond to a specific mineral grain of the specific portion (e.g., Fe2O3). Additionally, each new portion 401 may include a new second scan location 402, where the new second scan location 402 corresponds to the centroid or another estimate of the new portion 401.

[0431] The method is illustrated in Figure 5. A system may be configured to perform the method.

[0432] The method of FIG. 5 may include an image providing step S1, a first segmentation step S2, a first detection step S3, a first spectrum providing step S4, a first spectrum analysis step S5, and a classification step S6.

[0433] The image providing step S1 may include providing at least one image 300 of the sample 108 and / or a section thereof. The at least one image 300 may correspond to a backscattered electron image.

[0434] The first segmentation step S2 may include generating at least one or more second scan locations for at least one or more regions (see the description of FIG. 6 below). Each second scan location may be generated during the segmentation period. The method may further include storing coordinates of the second scan locations in a first data set 500. Each first data set may include coordinates of the second scan locations of one of the regions of the at least one image. In the example of FIG. 5, the method includes providing one first data set 500, which includes coordinates of five second scan locations represented by dots. The first detection step S3 may include detecting second emissions from the second scan locations 304 of the first data set 500. In particular, the first detection step may include scanning the electron beam across the second scan locations 304 and sequentially detecting corresponding second emissions. In this example, the second emissions include emissions of X-ray photons. The integration time for each second scanning location may correspond to a second dwell period. The method may further include performing the first segmentation step and the first detection step in parallel (if two or more regions of at least one image are processed). Furthermore, the method may include sequentially storing the detected X-ray photons for each second scanning location. The second dwell period may be four orders of magnitude longer than the first dwell period (see the description of FIG. 1). For example, the second dwell period may correspond to 8 milliseconds, and the first dwell period may correspond to 1 microsecond. Thus, the first and second emissions may be integrated over different dwell periods and detected over different scanning locations. Furthermore, the first and second emissions may be detected at different times. Thus, the image providing step and the first detection step may similarly occur at different times.

[0435] The first spectrum providing step S4 may include providing at least one or more first spectra 305, each generated based on the second emission detected at each of the second scanning locations 304 of at least one region 301 (i.e., the first data set). In the example of FIG. 5, the first spectrum providing step includes providing five first spectra 305 (one first spectrum per second scanning location) resulting from second scanning locations of the same first data set (i.e., region). However, the method may also include providing first spectra associated with multiple first data sets (i.e., regions). The first spectra 305 may correspond to a first X-ray spectrum. The first spectrum analyzing step S5 includes sequentially analyzing each first spectrum 305. Analyzing each first spectrum 305 may include comparing the respective first spectrum 305 with at least one or more reference spectra. Each reference spectrum may include a plurality of predetermined spectral lines associated with known mineral grains. The first spectral analysis step may further include a first line assignment step, which may include assigning spectral lines of each first spectrum 305 to predetermined spectral lines of a reference spectrum. Thus, the first spectral analysis step may include matching each first spectrum to one of the known mineral grains based on the results of the first line assignment step.

[0436] In practice, the first spectral analysis step includes calculating a confidence score for each first spectrum, where the confidence score corresponds to the level of match between the first spectrum and the corresponding matched reference spectrum. The confidence score may range from 0 to a maximum of 1. For example, a low confidence score of less than 0.95 may correspond to partial identification of the chemical composition of each mineral grain based on the corresponding first spectrum. Mineral grains with low confidence scores have a similar or nearly identical chemical composition to at least another mineral grain. Furthermore, the first spectral analysis step may include generating second datasets 501, where each second dataset 501 may include a list of coordinates of second scan locations associated with mineral grains with low confidence scores. In the example of FIG. 5, the method includes providing one second dataset including coordinates of three second scan locations 304, represented by dots, where each second scan location 304 is associated with a mineral grain requiring further processing for reliable classification. The method may further include providing the second dataset 501 as input to the classification step S5.

[0437] The classification step S5 may include a one-pass and / or two-pass classification step (two separate blocks) based on the results of the first spectral analysis step. The method includes performing a one-pass classification process on a second scanning location associated with at least one mineral grain having a similar chemical composition to the mineral grain. The one-pass classification step includes a second detection step, which includes detecting a second emission from the second scanning location in the second dataset 501 for the duration of a third dwell period. The third dwell period may be longer than the second dwell period of the first detection step S3. For example, the third dwell period may correspond to 80 milliseconds, and the second dwell period may correspond to 8 milliseconds. The third dwell period may depend on the corresponding calculated reliability score and the intrinsic photon count rate associated with each mineral grain. Therefore, the third dwell period may vary for each mineral grain. The method further includes performing a two-pass classification process on a second scanned location associated with specific mineral grains, where at least two of the specific mineral grains may contain the same or similar intensities on at least one image (e.g., Fe2O3 and Fe3O4). The two-step classification may include a second segmentation step and a separate second detection step. Both classification steps involve selecting mineral grains associated with initial low confidence scores and identifying them with high accuracy (e.g., 95% or greater). Figures 7a and 7b show in detail the steps comprising the one-pass and two-pass classification processes, respectively.

[0438] 6 shows a first flow chart of an alternative embodiment of this method. A system may be configured to perform the method.

[0439] The method may include step 600, which includes generating and providing at least one or more images, each corresponding to a backscattered electron image. Furthermore, each image may depict a sample or a section thereof. The method may also include step 601, which includes dividing at least one image into multiple sub-images using a contour algorithm. Each sub-image includes at least one image region 301, each corresponding to a grain within the sample. The method may also include step 602, which includes selecting and processing one of the sub-images. Processing one of the sub-images includes identifying portions within the respective region, each portion corresponding to a mineral grain of a corresponding grain. The method may also include step 603, which may include generating one second scan location for each portion using a k-means clustering algorithm and a flood-fill algorithm. Step 603 further includes generating each second scan location within a segmentation period. The method may further include step 606, which includes storing coordinates of the second scan locations of each sub-image in the first data set as an intermediate result. The method may also include step 607, which includes receiving notification that the first data set is available. Next, the method may perform step 608, which may include receiving the first data set and selecting one or more of the second scan locations of each first data set. The method may also include step 609, which may include sequentially detecting X-ray emissions from the second scan locations of each first data set for the duration of the second dwell period. Furthermore, the method may include decision block 610, which may include determining whether all first data sets have been selected for X-ray acquisition. If not, the method may include selecting another first data set in steps 607 and 608.Furthermore, the method may also include a decision block 604, which may include determining whether all subimages of at least one image have been selected and processed. If not, the method may include selecting another subimage in step 602. If all subimages have been selected and processed, the method performs step 605, which may include sending a notification to decision block 610, which may include marking the end of the segmentation processing for the corresponding image. In this case, decision block 610 determines that second scan locations from all first data sets have been selected for X-ray acquisition. The method may also include step 611, which may include generating a first X-ray spectrum based on the X-rays detected at each second scan location in step 609. The method further includes step 612, which may include storing the first X-ray spectrum of the corresponding second scan location in another data set as an intermediate result. The other data sets may include first X-ray spectra associated with one or more sub-images (i.e., regions) of the corresponding image. The method may also include step 613, which may include receiving notification that the other data sets are available. Next, the method may perform step 614, which may include receiving the other data sets and selecting one or more of the first X-ray spectra of each other data set. The method may also include step 615, which may include sequentially analyzing each first X-ray spectrum within each other data set. The analysis of the first X-ray spectra may include assigning spectral lines of each first X-ray spectrum to predetermined spectral lines of a reference spectrum, where each reference spectrum is associated with the X-ray spectrum of a known mineral grain. Step 615 may further include matching each first spectrum to one of the known mineral grains based on the results of the first line assignment step.Thus, step 615 may include calculating a confidence score for each first spectrum, where the confidence score corresponds to a level of match between the first X-ray spectrum and the corresponding matched reference spectrum. In other words, the confidence score may correspond to the accuracy of identification of the chemical composition (e.g., mineral composition) of each portion of the at least one image.

[0440] Further, the method may include a decision block 616, which may include determining whether the corresponding first X-ray spectrum requires further classification based on the confidence score calculated in step 615. If not, the method may include step 617, which may include terminating the classification of the corresponding first X-ray spectrum. The method may further include step 618, which may include storing the second scan location associated with a low confidence score in a second data set. A low confidence score may correspond to an identification accuracy of less than 95%. Further, the method may include a decision block 620, which may include determining whether all other data sets (of corresponding images) have been selected for analysis of the first X-ray spectrum. If not, the method may include selecting additional other data sets in steps 613 and 614. If decision block 610 determines that the first X-ray acquisition has been performed on the second scan locations of all of the first data sets of the corresponding images, the method may perform step 619, which may include sending a notification to decision block 620. Decision block 620 may then include determining whether to apply step 621a or step 621b, where step 621a includes a one-pass classification process and step 621b includes a two-pass classification process.

[0441] The first segmentation step may include steps 602, 603, and 604. The first detection step may include step 609. Steps 606 and 608 may include establishing a balancing queue of second scanning locations between the first segmentation step and the first detection step. Furthermore, the method may include performing the first segmentation step and the first detection step in parallel. Thus, step 609 may be performed in parallel with steps 602, 603, and 604. Step 615 may comprise a first spectral analysis step. The method may also include performing the first detection step and the first spectral analysis step in parallel.

[0442] The method may also include a decision block 622, which may include determining whether the number of images generated and provided is sufficient. If not, the method may include generating and providing another image at step 600. Alternatively, the method may include a step 623, which may include sending a notification to classification processes 621a and / or 621b, where the notification may include marking the end of image generation. More information on this may be found in the description of Figures 7a and 7b.

[0443] The method may also include storing the results of the corresponding classification process (step 621a or 621b) in step 624. The overall process ends in step 625.

[0444] FIG. 7a shows a second flowchart depicting the steps comprised by the one-pass classification process (620a).

[0445] The one-pass classification process may include step 700a, which may include receiving notification that a second data set of images is available for second X-ray detection. The one-pass classification process may also include sequentially processing the second data sets. Next, the method may perform step 701a, which may include receiving the second data sets and selecting one or more second scan locations for each second data set. The method may also include step 702a, which may include sequentially detecting X-ray emissions from the second scan locations of the corresponding second data sets for the duration of a third dwell period. The one-pass classification step may further include determining the third dwell period based on the reliability scores of each second scan location determined in the first spectral analysis step (step 614) and based on the characteristic photon count rate of each mineral grain. The third dwell period may be longer than the second dwell period of the first X-ray detection (step 609). For example, the third dwell period may correspond to 80 milliseconds, and the second dwell period may correspond to 8 milliseconds. Thus, step 702a may include detecting more X-ray photons within the third dwell period relative to the second dwell period. Furthermore, the one-pass classification step may include adding the number of X-ray photons detected by the second X-ray detection (step 702a) to the number of X-ray photons detected by the first X-ray detection (step 609) at each second scanning location of the second data set. Furthermore, the one-pass classification process may include a decision block 703a, which may include determining whether all second scanning locations of the second data set have been selected and further processed. If not, the method may include selecting another second data set in steps 700a and 701a.

[0446] The one-pass classification step may also include step 704a, which may include generating and storing, as an intermediate result, second X-ray spectra for the corresponding selected second scan locations in a new data set (i.e., a group of second spectra). Indeed, each new data set may include second X-ray spectra corresponding to the second scan locations obtained from at least one or more images. For example, the method may include acquiring a grid of images, each grid including 8x8 images. Each image within each grid may depict a section of the sample, and thus the grid of images may depict adjacent sections of the sample. Then, the new data set may include second X-ray spectra acquired from all images (8x8 images) within the corresponding grid.

[0447] When storing second X-ray spectra obtained from two or more images, the one-pass classification step may further include step 705a, which may include correcting image generation errors. In particular, the errors may include generating at least two images of adjacent sections of the sample, where the at least two images include at least two portions belonging to one of the mineral grains with a low confidence score. In other words, corresponding mineral grains are depicted in both images. Therefore, step 705a may include stitching together portions of the at least two images and merging the second X-ray spectra of each portion belonging to the same mineral grain with a low confidence score. The one-pass classification step may also include step 706a, which may include receiving notification of a new dataset (i.e., a group) that is available. Next, the method may perform step 707a, which may include receiving the new dataset and selecting one or more second X-ray spectra for each new dataset.

[0448] Furthermore, the one-pass classification step may also include step 708a, which may include sequentially analyzing each second X-ray spectrum of the new data set. The analysis of the second X-ray spectrum may include matching each second X-ray spectrum to at least two known mineral grains based on the results of the second line assignments. Therefore, step 708a may include calculating at least two new confidence scores and then selecting the highest one of the new confidence scores. The highest new confidence score for all second X-ray spectra of the new data set may correspond to a high confidence score of 95% or higher. In other words, the one-pass classification step may include calculating a high identification accuracy for the second spectra and thus normalizing the data quality for the second scan locations associated with mineral grains with (initial) low confidence scores. In fact, the one-pass classification process may reduce overall acquisition time by identifying unambiguous minerals with a minimum number of photons, while minerals with increasingly similar compositions may acquire additional photons, but only enough to resolve internal ambiguities between their subsets. Therefore, the one-pass classification step can be several times faster, at least five times faster, and preferably at least ten times faster, than conventional classification processes.

[0449] Step 702a may correspond to a second X-ray detection step, and step 708a may comprise a second spectral analysis step. When the first detection step (step 609) is completed, the method includes initiating a second detection step (step 702a) and completing a first spectral analysis (step 614) for the remaining second scan locations of the other data set. Thus, the method may include performing the second detection step (step 702a) and the first spectral analysis step (step 614) in parallel. Furthermore, the one-pass classification step includes performing the second spectral analysis step (step 708a) after the first spectral analysis step (step 614) has been fully performed for at least one or more images (e.g., 8x8 images). The method may also include performing the second detection step and the second spectral analysis step in parallel. In other words, the one-pass classification process may include performing a second spectral analysis on a first new data set of second spectra associated with a first grid of images (8x8 image) while concurrently performing a second detection step on second scan locations associated with a second grid of images (8x8 image), where the images are processed sequentially. The method may further include a decision block 710a, which may determine whether second X-ray spectra of all new data sets have been selected and analyzed. If not, the method may include selecting another new data set in step 706a.

[0450] As described above, the method performs step 623, which includes sending a notification to decision block 703a, which includes marking the end of image generation and therefore the end of the available second data set. In this case, the method includes step 709a, which includes sending a notification to decision block 710a, which includes the end of the available second X-ray spectrum (i.e., the available new data set).

[0451] FIG. 7b shows a third flowchart depicting the steps comprised by the two-pass classification process (620b). The two-pass classification process may include step 700b, which may include receiving a second data set of the image as input for the second segmentation step. Next, the method may perform step 701b, which may include sequentially selecting one or more of the second scan locations of each second data set. The two-pass classification process may further include step 702b, which may include sequentially redetecting the first emission from the first scan location of the (specific) portion 303 associated with the selected second scan location. Furthermore, step 702b may include reacquiring the first emission with adjusted contrast and brightness values. Each (specific) portion 303 (see description of FIG. 4 ) may correspond to a specific mineral grain (e.g., Fe2O3), which may include the same or similar intensity as at least one other specific mineral grain (e.g., Fe3O4) on the image. Next, the method may perform step 703b, which may include providing a new sub-image 400, which may include / delimit the corresponding specific portion 303 with adjusted contrast and brightness values. Furthermore, the two-pass classification process may include step 704b, which includes selecting and processing a corresponding new sub-image. Processing one of the new sub-images includes identifying a new portion 401 with increased contrast and brightness values. In fact, each of the new portions 401 may be a section of the initial specific portion 303. Thus, the two-pass classification process includes distinguishing at least two new portions 401 within each new sub-image based on their different intensities (e.g., gray-level intensities). The new portions 401 may correspond to two mineral grains (e.g., Fe2O3 and Fe3O4), and one of the mineral grains may correspond to the initially selected specific mineral grain (Fe2O3). The method may also include step 705b, which may include generating one new second scan location within another segmentation period for each new portion 401 using a k-means clustering algorithm and a flood-fill algorithm.

[0452] The method may further include step 707b, which includes storing coordinates of new second scan locations of each new sub-image in a third data set as an intermediate result. The method may also include step 708b, which includes receiving notification that the third data set is available for another second detection (i.e., X-ray detection). The method may also include step 709b, which may include receiving the third data set and sequentially selecting one or more of the new second scan locations of each third data set. Next, the method may perform another second X-ray acquisition in step 710b, which may include detecting X-ray emissions from the new second scan locations of each third data set (again sequentially) for the duration of another third dwell period.

[0453] Further, the method may include a decision block 712b, which may include determining whether all of the third data sets have been selected for another second x-ray acquisition. If not, the method may include selecting another third data set in steps 708b and 709b. Further, the method may also include a decision block 706b, which may include determining whether all new sub-images have been selected and processed. If not, the method may include selecting another new sub-image in step 704b. As mentioned in the description of FIG. 6, the method may include step 623, which may include sending a notification to decision block 706b, which may include marking the end of image generation and, therefore, the end of providing new sub-images. Thus, when all available new sub-images have been selected and processed, the method performs step 711b, which includes sending a notification to decision block 712b, which includes marking the end of the (second) segmentation process for the new sub-images. In this case, decision block 712b determines that new second scan locations from all third data sets have been selected for the other second X-ray acquisition. The second segmentation step may include steps 704b, 705b, and 706b. The other second detection step may include step 710b. Steps 707b and 709b may include establishing a balancing queue for the new second scan locations between the second segmentation step and the other second detection step. Furthermore, the two-pass classification process may include performing the second segmentation step and the other second detection step in parallel.

[0454] Once the first detection step (step 609) is completed, the method includes commencing a second segmentation step and completing a first spectral analysis step (step 614) for the remaining second scan locations of the other data set. Thus, the method may include performing the second segmentation step and the first spectral analysis step (step 614) in parallel. The method may also include performing other second detection steps and first spectral analysis steps (step 614) in parallel (until the first spectral analysis step is completed).

[0455] The two-pass classification step may also include step 713b, which may include generating and storing, as an intermediate result, a new X-ray spectrum for a corresponding new second scan location in another new data set (i.e., another group). As described above, each other new data set may include a new X-ray spectrum corresponding to the new second scan location obtained from at least one or more images. The two-pass classification step may also include step 714b, which may include correcting image generation errors (see the description of FIG. 7a above). Furthermore, the method may include step 715b, which may include receiving notification that another new data set is available. Next, the method may perform step 716b, which may include receiving another new data set and selecting (again sequentially) one or more new X-ray spectra for each other new data set.

[0456] Furthermore, the two-pass classification step may also include performing another second spectral analysis in step 717b, which may include sequentially analyzing each new X-ray spectrum of the other new data set. The analysis of the new X-ray spectra may include matching each new X-ray spectrum to at least two known mineral grains based on the results of the other second line assignment (as described above). Thus, step 717b may include calculating at least two or more other new confidence scores and then selecting the highest of the other new confidence scores. The highest new confidence score for all new X-ray spectra of the new data set may correspond to a high confidence score of 95% or greater.

[0457] Step 717b may be configured with another second spectral analysis step. The method may also include performing another second detection step and another second spectral analysis step in parallel. Furthermore, the two-pass classification step includes performing the second spectral analysis step (step 717b) after the first spectral analysis step (step 614) has essentially been performed on at least one or more images (e.g., 8x8 images).

[0458] The method may further include a decision block 719b that may determine whether all of the other new data sets (i.e., other groups) have been selected and analyzed. If not, the method may include selecting additional other new data sets in steps 715b and 716b. If new second scan locations for all third data sets have been selected and processed, the method performs step 718b, which may include sending a notification to decision block 719b, where the notification may include marking the end of the other second detection step.

[0459] The two-pass classification process involves identifying the chemical composition of the new moieties 401 in each respective new sub-image based on the results of another second spectral analysis step and a second segmentation step.

[0460] A system is shown in Figure 8. The system may be configured to perform the method.

[0461] The system includes a scanning microscope system 100 and a data processing system 800 .

[0462] The data processing system 800 may include one or more processing units configured to execute computer instructions (i.e., machine-readable, executable instructions) of a program. The processing unit may be singular or multiple. For example, the data processing system 800 may include at least one of a CPU, a GPU, a DSP, an APU, an ASIC, an ASIP, or an FPGA. In this example, the processing unit may be configured to form an X-ray spectrum based on the detected X-rays. In particular, for the EDS modality, the processing unit may be configured to count and sort the detected X-rays (for each second scan location) based on the energy of each X-ray over the duration of the second dwell period. However, for the WDS modality, the processing unit may be configured to count and sort the detected X-rays based on the wavelength of each X-ray during the second dwell period.

[0463] Data processing system 800 may include memory components, such as data storage component 810. Data storage component 810, as well as data processing system 800, may include at least one of main memory (e.g., RAM), cache memory (e.g., SRAM), and / or secondary memory (e.g., HDD, SDD).

[0464] Data processing system 800 may include volatile and / or non-volatile memory, such as SDRAM, DRAM, SRAM, flash memory, MRAM, F-RAM, or P-RAM. Data processing system 800 may include an internal communication interface (e.g., a bus) configured to facilitate electronic data exchange between components of data processing system 800, such as communication between memory components and processing components.

[0465] Data processing system 800 may include an external communication interface configured to facilitate electronic data exchange between the data processing system and a device or network external to the data processing system. In the example of Figure 8, the external communication interface may be configured to facilitate electronic connections between processing components of data processing system 800 and components of scanning microscope system 100, such as control unit 115. Additionally, the external communication interface may be configured to establish electronic data exchange between processing components of data processing system 800 and vacuum controller 116.

[0466] Furthermore, the external communication interface may also be configured to establish electronic data exchange between the data processing system 800 and the first detector 111. The external communication interface may also be configured to facilitate an electronic connection between the data processing system 800 and the second detector 112. For example, the detected backscattered electron data from all the first scan locations may be stored in the data storage component 810. The processing unit of the data processing system 800 may be configured to form at least one image based on the stored backscattered electron data.

[0467] The backscattered electron image of the sample and the x-ray spectrum from each second scan location may be stored in data storage component 810.

[0468] The data processing system may also include a network interface card that may be configured to connect the data processing system to a network, such as the Internet. The data processing system may be configured to transfer electronic data using a standardized communication protocol. The data processing system may be a centralized or a distributed computing system.

[0469] The data processing system may include a user interface, such as an output user interface and / or an input user interface. For example, the output user interface may include a screen and / or monitor configured to display visual data (e.g., a backscattered electron image or an X-ray spectrum of the sample) or a speaker configured to communicate audio data (e.g., to play audio data to a user). The input user interface may be, for example, a keyboard configured to allow insertion of text and / or other keyboard commands (e.g., allowing a user to enter instructions or method parameters into the scanning microscope system) and / or a trackpad, mouse, touchscreen, and / or joystick configured to navigate, for example, the backscattered electron image or an identified region in the backscattered electron image.

[0470] Briefly, data processing system 800 may be a processing unit configured to execute instructions of a program. Data processing system 800 may be a system-on-chip including a processing unit, memory components, and a bus. Data processing system 800 may be a personal computer, a laptop, a pocket computer, a smartphone, or a tablet computer. The data processing system may comprise a server, a server system, part of a cloud computing system, or a system emulating a server, such as a server system with appropriate software for running virtual machines. The data processing system may be a personal computer, a laptop, a pocket computer, a smartphone, a tablet computer, and / or a processing unit or system-on-chip that may interface with a user interface (such as the user interfaces described above).

[0471] 8, the data processing system includes a one-pass classification component 860a configured to perform at least a portion of the one-pass classification step. The data processing system further includes a two-pass classification component 860b configured to perform at least a portion of the two-pass classification step.

[0472] In other words, data processing system 800 may include one-pass and two-pass classification components 850a and 850b, respectively. More specifically, data processing system 800 may include at least one storage device on which data processing system 800 may be stored.

[0473] At least one of the two classification components 850a and 850b may be implemented in software. Thus, at least one of the classification components 850a and 850b may be a software component, or at least a portion of one or more software components. The data processing system 800 may be configured to execute the software component and / or to execute software that includes the software component. In other words, at least one of the classification components 850a and 850b may include one or more computer instructions (i.e., machine-readable instructions) that can be executed by a computer (e.g., the data processing system 800).

[0474] The one-pass and / or two-pass classification components 850a and 850b may be stored on one or more different storage devices. For example, the classification components 850a and 850b may be stored on multiple storage devices, including persistent memory, e.g., multiple storage devices in a RAID system, or multiple storage components including different types of memory, such as persistent memory (e.g., HDD, SDD, flash memory) and main memory (e.g., RAM).

[0475] At least one of the classification components 850a and 850b may also be implemented at least partially in hardware. For example, the classification components 850a and 850b, or at least a portion of at least one of the classification components 850a and 850b, may be implemented as a programmed and / or customized processing unit, hardware accelerator, or system-on-chip that may interface with the data processing system 800, a personal computer, laptop, pocket computer, smartphone, tablet computer, and / or server.

[0476] At least one of the classification components 850a and 850b may also include hardware-implemented and software-implemented elements, as an example, there may be the use of a hardware-implemented encryption / decryption unit and software-implemented processing of the decrypted data.

[0477] The data processing system 800 may also include a first segmentation component 820. The first segmentation component 820 may be configured to perform the first segmentation step. More specifically, the data processing system 800 may include at least one storage device in which the first segmentation component 820 may be stored.

[0478] The data processing system 800 may also include a pre-processing component 830. The pre-processing component 830 may be configured to perform a pre-processing step.

[0479] The data processing system 800 may also include a first spectral analysis component 840. The first spectral analysis component 840 may be configured to perform a first spectral analysis step.

[0480] Additionally, data processing system 800 may also include a post-processing component 860. Post-processing component 860 may be configured to perform post-processing steps.

[0481] The data processing system 800 may include at least one storage device, such as a data storage component 810, in which at least one of the first segmentation component 820, the pre-processing component 830, the first spectral analysis component 840, and the post-processing component 860 may be stored.

[0482] At least one of the first segmentation component 820, the pre-processing component 830, the first spectral analysis component 840, and the post-processing component 860 may be implemented in software. One, some, or all of these components may be software components, or at least portions of one or more software components. The data processing system 800 may be configured to execute the software components and / or to execute software that includes the software components. In other words, the components may include one or more computer instructions (i.e., machine-readable instructions) that can be executed by a computer (e.g., the data processing system 800).

[0483] At least one of the first segmentation component 820, the pre-processing component 830, the first spectral analysis component 840, and the post-processing component 860 may be stored in one or more different storage devices. For example, at least one of the components may be stored in multiple storage components including persistent memory, e.g., multiple storage devices in a RAID system, or different types of memory, such as persistent memory (e.g., HDD, SDD, flash memory) and main memory (e.g., RAM).

[0484] The components may also be implemented at least partially in hardware. For example, at least one of the first segmentation component 820, the pre-processing component 830, the first spectral analysis component 840, and the post-processing component 860, or part of the functionality of one of those components, may be implemented as a programmed and / or customized processing unit, hardware accelerator, or system-on-chip that may interface with the data processing system 800, a personal computer, a laptop, a pocket computer, a smartphone, a tablet computer, and / or a server.

[0485] Although a preferred embodiment has been described above with reference to the accompanying drawings, those skilled in the art will understand that this embodiment is provided for illustrative purposes only and should not be construed in any way as limiting the scope of the present invention as defined by the claims.

[0486] Whenever relative terms such as "about," "substantially," or "approximately" are used herein, such terms should also be construed to include the exact terms. That is, for example, "substantially straight" should also be construed to include "(exactly) straight."

[0487] Whenever steps are recited above or in the appended claims, it should be noted that the order in which the steps are recited in this text may be incidental. That is, unless otherwise specified or apparent to one of ordinary skill in the art, the order in which the steps are recited may be incidental. That is, if the specification describes, for example, that a method includes steps (A) and (B), this does not necessarily mean that step (A) precedes step (B); step (A) may be performed (at least partially) simultaneously with step (B), or step (B) may precede step (A). Furthermore, if step (X) is described as preceding another step (Z), this does not mean that there are no steps between step (X) and step (Z). That is, step (X) preceding step (Z) encompasses the situation in which step (X) is performed immediately before step (Z), but also encompasses the situation in which step (X) is performed before one or more steps (Y1), ..., that follow step (Z). Corresponding considerations apply when terms such as "after" or "before" are used.

[0488] Reference symbol 100 Scanning Microscope System 101 Scanning Electron Microscope 102 Electron source 103 Anode 104 Condenser Lens 105 scanning coil 106 Objective Lens 107 Electron Beam 108 samples 109 First Release 110 Second Release 111 First detector 112 Second Detector 113 Movable Stage 114 Centerline of second detector 115 Control Unit 116 Vacuum Controller 117 Mechanical Pumping System 118 Ultra-high vacuum pump 119 Vacuum chamber 200 particles 201 Epoxy Matrix 202 First Scan Location 203 dashed line 300 images 301 area 302 sub-images 303 (specific) part of a sub-image 304 Second Scan Location 305 Spectrum 306 Spectral Lines 400 new sub-images 401 New part of new sub-image 402 New Second Scan Location S1 Image provision step S2 First detection step S3 First spectrum provision step S4 First spectral analysis step S5 Classification step 500 First Dataset 501 Second Dataset 600-625 Steps of the embodiment of the method shown in the flowchart of FIG. 700a-710a Steps of the one-pass classification process (step 621a in FIG. 6) shown in the flowchart of FIG. 7a 700b-719b Steps of the two-pass classification process (step 621b in FIG. 6) shown in the flowchart of FIG. 7b 800 Data Processing Systems 810 Data Storage Component 820 First Segmentation Component 830 Pretreatment Components 840 First Spectral Analysis Component 850a 1-pass classification components 850b Two-pass classification components 860 Aftertreatment Components

Claims

1. 1. A system configured for materials analysis and mineralogy, comprising: a scanning microscope system having a first detector and a second detector; and a data processing system having a data storage component and a first spectral analysis component; the data storage component is configured to provide at least one image of the sample or section thereof based on first emissions detected by the first detector within a first dwell period from a plurality of first scan locations; the second detector is configured to detect second emissions from at least one second scan location of at least one region of the at least one image during a second dwell period, each second scan location associated with a portion of a corresponding region; the data storage component is configured to provide at least one first spectrum, each first spectrum based on the second emissions detected at each of the second scan locations of the at least one region; the first spectrum analysis component is configured to calculate a reliability score for all of the first spectra and to select the second scan locations associated with the first spectra whose reliability scores are below a threshold; the second detector is configured to detect the second emission from at least one of the selected second scan locations for a third dwell period, and the data storage component is configured to provide at least one new image delimiting a portion associated with the selected second scan location, and to determine a new second scan location in each new image via its modified contrast and brightness values ​​for the at least one image; the data processing system includes a first segmentation component configured to determine the second scanning location of regions of the at least one image, the regions corresponding to grains within the sample and portions within the regions corresponding to mineral grains within each grain; the third dwell period is longer than the second dwell period; the third dwell period is dependent on the corresponding calculated confidence score and the unique count rate associated with each mineral grain; system.

2. 2. The system of claim 1, wherein the scanning microscope system is configured to generate the first emission and the second emission, the first emission comprising an emission of particles (e.g., backscattered electrons) and the second emission comprising an emission of photons (e.g., X-ray photons).

3. 3. The system of claim 1, wherein the data processing system is configured to generate at least one image and a new image based on the first emission detected at each first scanning location, and / or the at least one image corresponds to a backscattered electron image, and / or the new image corresponds to a backscattered electron image.

4. 4. The system of claim 1, wherein the data processing system is configured to generate the first spectra based on the second emissions detected at each of the second scan locations of the region, each first spectrum corresponding to an X-ray spectrum.

5. 5. The system of claim 1, wherein the confidence score below the threshold corresponds to a low confidence score, the low confidence score corresponding to a partial identification of the chemical composition of the respective mineral grain based on the corresponding first spectrum, and the mineral grain with the low confidence score comprises a chemical composition similar to at least another mineral grain.

6. 6. The system of claim 1, wherein the data processing system is configured to generate at least one second spectrum, each second spectrum including a total number of photons (e.g., X-ray photons) detected at a corresponding selected second scan location during the second dwell period and the third dwell period.

7. 6. The system of claim 1, wherein the data processing system is configured to calculate at least two new reliability scores for each second spectrum, and wherein the highest new reliability scores for at least some of the second spectra correspond to a high reliability score (above a threshold).

8. 8. The system of claim 1, wherein the data processing system is configured to use the adjusted contrast and brightness values ​​of each new image to reveal and / or detect at least two new portions in at least one of the new images, the new portions corresponding to mineral grains that include the same or similar intensities on the at least one image (i.e., indistinguishable mineral grains).

9. 9. The system of claim 1, wherein the second detector is configured to detect the second emission from the new second scan location of the new portion for the duration of another third dwell period.

10. 1. A method for determining a property of a sample or a section thereof, comprising: providing at least one image of said sample or section thereof based on first emissions detected within a first dwell period from a plurality of first scanning locations; - performing a first detection step, said first detection step comprising detecting second emissions during a second dwell period from at least one second scanning location of at least one region of at least one image, each second scanning location being associated with a portion of a corresponding region; - performing a first spectrum providing step, said first spectrum providing step comprising providing at least one first spectrum, each first spectrum being based on the second emissions detected at each of said second scan locations of said at least one region; - performing a first spectral analysis step, the first spectral analysis step comprising calculating a reliability score for all first spectra and selecting the second scan locations associated with the first spectra having a reliability score below a threshold; performing a classification step, the classification step comprising detecting the second emission from at least one of the selected second scan locations during a third dwell period and providing at least one new image delimiting a portion associated with the selected second scan location, and determining a new second scan location in the corresponding new image via its modified contrast and brightness values ​​for the at least one image; the method further comprising a first segmentation step, the first segmentation step comprising determining the second scanning location of regions of the at least one image, the regions corresponding to grains within the sample and portions within the regions corresponding to mineral grains within each grain; the third dwell period is longer than the second dwell period; the third dwell period is dependent on the corresponding calculated confidence score and the unique count rate associated with each mineral grain; method.

11. 11. The method of claim 10, wherein the method comprises generating the first emission and the second emission, the first emission comprising an emission of particles (e.g., backscattered electrons) and the second emission comprising an emission of photons (e.g., X-ray photons).

12. 12. The method of claim 10 or 11, comprising generating at least one image and a new image based on the first emission detected at each first scanning location, and / or the at least one image corresponds to a backscattered electron image, and / or the new image corresponds to a backscattered electron image.

13. 13. The method of claim 10, further comprising generating the first spectra based on the second emissions detected at each of the second scan locations of the region, each first spectrum corresponding to an X-ray spectrum.

14. 14. The method of claim 10, wherein a confidence score below the threshold corresponds to a low confidence score, the low confidence score corresponding to a partial identification of the chemical composition of the respective mineral grain based on the corresponding first spectrum, and the mineral grain with the low confidence score comprises a chemical composition similar to at least another mineral grain.

15. 15. The method of claim 10, wherein the classifying step comprises a one-pass classifying step, the one-pass classifying step comprising generating at least one second spectrum, each second spectrum comprising a total number of photons (e.g., X-ray photons) detected during the second dwell period and the third dwell period at a corresponding selected second scan location.

16. A method described in any one of claims 10 to 15, wherein the classification step includes a one-pass classification step, the one-pass classification step including calculating at least two new reliability scores for each second spectrum, and the highest new reliability score for at least some of the second spectra corresponds to a high reliability score (above a threshold).

17. 17. A method according to any of claims 10 to 16, wherein the classification step comprises a two-pass classification step, the two-pass classification step comprising using the adjusted contrast and brightness values ​​of each new image to reveal and / or detect at least two new portions in at least one of the new images, the new portions corresponding to mineral grains comprising the same or similar intensity on at least one image (i.e. indistinguishable mineral grains).

18. A method according to any one of claims 10 to 17, wherein the classification step includes a two-pass classification step, and the two-pass classification step includes detecting the second emission from the new second scanning location of the new portion for the duration of another third dwell period.

19. A computer program comprising instructions which, when executed by a computer, cause the computer to perform the steps of the method according to any one of claims 10 to 18.

Images