Systems and methods for pixel detection

Abstract

Aspects of the present technology provide improved pixel detection techniques, including improved motion detection and processing resource savings. [Solution] The improved technique includes determining a mutual information metric between pairs of images from an image sequence of a biological specimen, detecting motion of the biological specimen based on the mutual information metric, and performing measurements of the biological specimen based on test images of the image sequence after motion is detected.

Description

[Technical Field]

[0001] The present invention relates generally to image processing and biological classification and measurement. [Background technology]

[0002] Pixel detection techniques generally involve analyzing or measuring biological specimens based on captured digital images of the biological specimen, e.g., pixel-based images. For example, the biological specimen may be mounted in a microscope capable of capturing digital images or videos, and the resulting digital images may be analyzed to classify or measure the biological specimen. However, existing techniques exhibit certain drawbacks, such as relatively high CPU utilization. Therefore, there is a need in the art for improved pixel detection techniques. Summary of the Invention

[0003] To meet the described long-standing needs, the present disclosure provides an image processing method that includes determining a mutual information metric between at least one pair of images from an image sequence of a biological specimen, detecting motion of the biological specimen based on the mutual information metric, and performing a measurement of the biological specimen based on a test image of the image sequence after motion is detected.

[0004] Also provided is an image processing method that includes determining a mutual information metric between at least one pair of images from an image sequence of a biological sample, detecting motion of the biological sample based on the mutual information metric, and performing a measurement of the biological sample based on a test image of the image sequence after motion is detected.

[0005] Further provided is a non-transitory computer-readable memory storing instructions that, when executed by the processor, cause the processor to determine a mutual information metric between at least one pair of images from a sequence of images of the biological specimen, detect movement of the biological specimen based on the mutual information metric, and, after movement is detected, perform a measurement of the biological specimen based on a test image of the image sequence. [Brief explanation of the drawings]

[0006] Particular features of the present technology are set forth in the appended claims, however, for purposes of explanation, some implementations of the present technology are set forth in the following exemplary, non-limiting figures. [Figure 1] 1 illustrates an exemplary image processing scenario. [Figure 2] 1 illustrates an exemplary image processing system in accordance with aspects of the present technique; [Figure 3] 1 illustrates an exemplary pixel detection system in accordance with aspects of the present technique; [Figure 4] 1 illustrates an exemplary mutual information measurement system in accordance with aspects of the present technique. [Figure 5] 1 illustrates an exemplary method for biological measurements in accordance with aspects of the present technology. [Figure 6] 1 illustrates an exemplary computing device in accordance with aspects of the present technology. DETAILED DESCRIPTION OF THE INVENTION

[0007] The detailed description set forth below is intended as a description of various configurations of the present technology and is not intended to represent the only configurations in which the present technology may be practiced. The accompanying drawings are incorporated herein and constitute a part of the detailed description. The detailed description includes specific details for the purpose of providing a thorough understanding of the present technology. However, the present technology is not limited to the specific details described herein and can be practiced using one or more other implementations. In one or more implementations, structures and components are shown in block diagram form to avoid obscuring the concepts of the present technology.

[0008] The present disclosure provides, among other things, improved pixel detection techniques. Common image processing techniques often do not work well for laboratory images of biological specimens, and therefore techniques adapted to address such images may yield improved results. In particular, transmitted-light images, in which a primary light source and camera located behind the specimen capture light after it passes through the specimen rather than reflecting off it, and images of emissive specimens, in which the specimen generates and emits electromagnetic energy that is captured by the camera independently of any other light source, often do not work well with pixel detection or other image processing techniques designed for reflected-light images, in which a primary light source is reflected off the specimen. Furthermore, specimens with a substantial fluid component or submerged in a fluid medium can confuse or reduce the effectiveness of conventional image processing techniques, especially when captured in transmitted images.

[0009] In one aspect of the subject technology presented herein, mutual information can form the basis of improved techniques for identifying movement or other changes in biological specimens. For example, the mutual information metric between two consecutive images of a given image object, such as a biological specimen, can be used to detect or measure object movement or other changes that occurred between the capture times of the two images. Such movement can be, for example, cell swelling, cell contraction, or cell translation. Some examples of such other changes include, for example, an increase or decrease in the amount of a cellular component or cellular product, movement of cellular components within a cell, cell replication, etc. The mutual information metric can be based, for example, on a measure of statistical independence between two images or between corresponding pixels in two images. Experimental results have shown that the mutual information metric can provide improved detection or classification of movement or other changes in images of biological specimens. For example, the mutual information metric can be relatively more sensitive to changes in the biological specimen that are relevant to some clinical applications, such as cell growth or movement, but relatively less sensitive to other, less relevant changes, such as movement of the medium in which the biological specimen is submerged.

[0010] In another aspect of the disclosed technology, measures of motion or other changes in an image object, such as a biological specimen, can be used to improve the performance of biological measurements. For example, a resource-intensive task, such as performing a pixel detection method on a test image, or any other resource-intensive biological measurement, can be initiated when motion or other changes are detected in the biological specimen. In one aspect, resources, such as computer processor or memory usage, can be saved by forgoing the resource-intensive task when no changes are detected.

[0011] The improved pixel detection technique can include determining a mutual information metric between at least one pair of images from an image sequence of the biological specimen, detecting motion of the biological specimen based on the mutual information metric, and performing a measurement of the biological specimen, such as a density metric, based on a test image of the image sequence after motion is detected. Such an image can be, for example, a transmitted light image. In improved techniques, (i) the technique may include forgoing performing a measurement of the biological sample if motion is not detected; (ii) determining the mutual information metric may include estimating a measure of statistical independence between pixel values ​​at the same location in the image pair; (iii) estimating the measure of statistical independence may include determining a joint histogram of pixel values ​​at the same location; (iv) motion is detected when the mutual information metric exceeds a mutual information threshold level; (v) motion may be detected based on multiple metrics of mutual information, each of the multiple metrics being between a different pair of images from the sequence of images; (vi) performing the measurement of the biological sample may be delayed after motion is detected until motion is no longer detected; and / or (vii) performing the measurement of the biological sample may include processing test images with a machine learning model to generate the measurement of the biological sample.

[0012] In a further aspect of the improved technique, performing a measurement of the biological sample can include analyzing a test image from the image sequence to generate a plurality of feature images, deriving a likelihood image for each of a plurality of object types from the feature images, and integrating the likelihood images into a classification image indicating which of the plurality of object types is detected at each pixel in the classification image. An exemplary biological measurement can include determining a density metric, and the improved technique can include calculating the density metric for an object type based on the percentage of pixels in the classification image that indicate the object type.

[0013] 1 illustrates an exemplary image processing scenario 100. Scenario 100 includes a test strip 102 of cells 112 in a test strip container 110, a camera 104 configured to capture images of the test strip 102, a light source 114, an image processor 106, and a display 108. The light source 114 may emit visible light or other electromagnetic radiation, and the light source 114 may be positioned relative to the test strip 102 facing the camera 104 such that emission from the light source 114 can radiate in a direction 116 and pass through a translucent or transparent portion of the test strip 102 to be captured by a light sensor within the camera 104. The image processor 106 may process one or more images of the test strip 102 captured by the camera 104 to generate biological measurements of the test strip 102, and the resulting measurements, such as confluency measurements, may be presented to a user on the display 108. In one embodiment, the measurements presented on the display 108 are updated only when the image processor detects a particular type of change or movement of the test strip 102. For example, if the user repositions the test strip 102 relative to the camera 104 so that the camera 104 captures an image of a different portion of the test strip 102, the user may want the biological measurements to be updated immediately. At other times, for example, when the sample 102 is not being moved by the operator, the image processor 106 may forgo updating the biological measurements to conserve resources used by the image processor 106 when the biological measurements are unlikely to have changed.

[0014] FIG. 2 illustrates an exemplary image processing system 200 according to aspects of the present technology. System 200 may be one exemplary implementation of image processor 106 (FIG. 1). System 200 includes a motion detector 202, a controller 204, and a pixel detector 206. During operation, an image source may provide a sequence of images captured at different times. In one aspect, one or more images in the captured sequence may include the same image subject, such as images of the same biological sample. Motion detector 202 may evaluate motion or other changes occurring between pairs of images from the image source. In one aspect, the image pairs may be adjacent or consecutive images from the image source, while in another aspect, the image pairs may be further apart in time from each other and represent a greater difference in image capture time points. In one aspect, motion detector 202 may include a mutual information measurement 208, and motion detector 208 may detect motion based on the measured mutual information. For example, the mutual information metric can indicate the degree of motion that occurred between the capture times of the image pair, and the mutual information metric can be normalized as described below. In another example, motion can be detected when the mutual information metric falls below a threshold level. The pixel detector 206 can perform biological measurements of the image subject in a test image from the image source. In one aspect, the test image used by the pixel detector 206 can be one of the image pairs used by the motion detector 202, e.g., the most recent of the pair. In another aspect, the test image used by the pixel detector 206 can be different from the image in the image pair used by the motion detector 207, e.g., the test image can be newer, or more recent, than either image in the image pair. The controller 204 can control the performance of biological measurements by the pixel detector 206 based on the motion detected by the motion detector 202. For example, the controller 204 can initiate biological measurements by the pixel detector 206 only when the motion detector detects a certain amount or quality of motion between the image pair.

[0015] In some optional aspects of system 200, pixel detector 206 may include one or more of feature generator 210, likelihood generator 212, pixel classifier 213, density generator 214, and one or more machine learning models 216. In one aspect, the one or more machine learning models may analyze the test image to generate measurements of objects in the test image. In another aspect, feature generator 210, likelihood generator 212, pixel classifier 213, and density generator 214 may be used in combination to generate measurements of objects in the test image. In some implementations, feature generator 210, likelihood generator 212, pixel classifier 213, and / or density generator 214 may each individually include a machine learning model.

[0016] In another additional aspect, the pixel detector 206 can generate a biological measurement, such as the density of a specimen captured in a test image. For example, a density metric for the specimen can be determined by the density generator 214 from the output of the pixel classifier 214. The density metric can be determined as a ratio of counts of pixels having different classifications generated by the pixel classifier for the image. For example, the density metric can be determined as the ratio of a count of the number of pixels in the test image classified as a particular type of cell divided by the total number of pixels in the test image.

[0017] 3 illustrates an exemplary pixel detection system 300 in accordance with aspects of the present technique. The pixel detection system 300 may be one exemplary implementation of the pixel detector 206 (FIG. 2). The system 300 includes a feature generator 302, a likelihood generator 304, and a pixel classifier 306. In operation, the feature generator 302 can generate one or more feature images 312 from a test image 310. The likelihood generator can generate one or more likelihood images 314 from the feature images 312. The pixel classifier 306 can generate a classification image 316 based on the likelihood images 314.

[0018] The test image 310 may be an image from an image source such as the camera 104 (FIG. 1) and may be, for example, a color image with multiple color component values ​​per pixel or a grayscale image with a single color (grayscale) component per pixel. The feature image 312 may indicate the location of features in the test image 310, with pixel values ​​in the feature image indicating the presence of a feature type at that pixel location. Each feature image 312 generated from one test image 310 may correspond to a different feature type and have a different localization, such as a computer vision feature (e.g., edge, texture, motion, etc.) or a statistical feature (e.g., a spatially local mean or standard deviation of pixel intensity values). Feature images with different localizations may use different localization techniques to characterize the feature type, for example, by varying a window size around the output pixel to which the source pixel is considered local (e.g., varying the radius from the output pixel). For example, the feature detector 302 may generate six feature images 312 from one test image 310, including three mean images with localization radii of 2, 3, and 4 pixels, and two standard deviation images with localization radii of 2 and 5 pixels.

[0019] In some aspects, the feature generator 302 may apply convolution filters to the test image 310 to generate feature images 312 for features such as: Gaussian-weighted intensity features within local windows of various sizes to determine features at different scales, Gaussian-weighted local variance features with various window sizes, and / or Gabor filters to identify image pattern or texture features.

[0020] Each likelihood image 314 may indicate the likelihood of a corresponding object type that may be present in the image subject of test image 310. For example, each pixel value in a likelihood image may indicate an estimated probability that an object type, such as a particular organelle, is present at each pixel's corresponding location in test image 310. In some aspects, one or more of feature images 312 may be used by a likelihood generator to generate each likelihood image 314.

[0021] In one example, the likelihood generator 304 may generate a likelihood image for a particular classification class with an adjustable pseudo-sensitivity parameter s by calculating the pixel-by-pixel probability as follows:

[0022]

number

[0023] The probability that a pixel belongs to a certain class can be determined based on a multivariate statistical distribution model generated from manually annotated training images. Such a model can model each class as a normal distribution using a full covariance matrix. The training images can include manual annotations to distinguish between background and foreground pixel classes, or between background, cell edge, and cell center pixel classes.

[0024] In one aspect, the improved pixel detection techniques may include techniques for faster and / or more efficient processing, such as processing multivariate statistical models using single instruction multiple data (SIMD) parallel processing to improve pixel accuracy. 分類 By parallelizing the computation of , it is possible to improve performance during generation of the classified image 316.

[0025] The pixel classifier 306 may combine the likelihood image 314 into a classification image 316 to indicate which object type was detected, if any, at the corresponding pixel of the test image 310. The pixel classifier 306 may, for example, select which object type is most likely to be present at each pixel location. Alternatively, the pixel classifier 306 may indicate a count of objects detected at each pixel, or each pixel may indicate which combination of objects was detected at each pixel (e.g., using different colors to indicate the presence of different object types). In some embodiments, the pixel classifier may use a likelihood threshold to determine whether an object type is present at a pixel location.

[0026] 4 illustrates an exemplary mutual information measurement system 400 in accordance with aspects of the present technology. The system 400 may be one exemplary implementation of the mutual information measurement unit 208 (FIG. 2). The system 400 includes a statistics collection unit 404, an entropy estimation unit 406, and a mutual information metric calculation unit 408. During operation, the statistics collection unit 404 may collect statistics of pixel data in the image pair 402. The entropy estimation unit 406 may estimate entropy in the image pair 402 based on the statistics collected by the statistics collection unit 404. The mutual information metric calculation unit 408 may calculate a mutual information metric for the image pair 402 based on the entropy estimated by the entropy estimation unit 406.

[0027] In an optional aspect of system 400, statistics collection unit 404 may include a marginal histogram unit 410 and a joint histogram unit 412. Entropy estimation unit 406 may include a marginal entropy unit 414 and a joint entropy unit 416. In one aspect, the marginal entropy estimation in box 414 may be based on the marginal histogram from box 410, and the joint entropy estimation in box 416 may be based on the joint histogram from box 412. In one aspect, the joint histogram of image pair 402 may count the frequency of occurrence of pixel values ​​at the same location in the two images. For example, each entry in the joint histogram may indicate a count of the number of pixels in the image pair where one image of the image pair has a first pixel value and the corresponding pixel at the same location in the other image has a second pixel value.

[0028] In some implementations, the mutual information metric between two images may be normalized based on individual estimates of the entropy of each of the two images. For example, the normalized mutual information I 正規化 teeth,

number

[0029] 5 illustrates an exemplary method 500 for biological measurements in accordance with aspects of the present technology. Method 500 may be an exemplary method performed by image processor 106 (FIG. 1) or system 200 (FIG. 2). Method 500 includes calculating mutual information from image pairs (506), detecting motion based on the calculated mutual information (508), and performing biological measurements of image subjects in test images (box 512).

[0030] The mutual information metric may be based on Shannon's definitions of entropy H and mutual information I(X;Y). In operation of one exemplary implementation, the mutual information metric (506) may be based on collected statistics and entropy estimates, as described above with respect to the mutual information measurement system 400 (FIG. 4). In one aspect, statistics for image pairs may be collected (box 504), as described above with respect to FIG. 4, and these statistics may be based on a selected bin size for the histogram (502).

[0031] Motion may be detected (508), for example, when the mutual information falls below a threshold level, indicating that one image of the pair is not sufficiently predicted by the other image of the pair. In one embodiment, performance of the biological measurement (512) may be delayed (510) for a period of time following the initial detection of motion (508). For example, after the initial detection of motion, the biological measurement (512) may not begin until after motion is no longer detected. In one embodiment, a first threshold of the mutual information metric may be used to detect when motion begins, while a second threshold of the mutual information metric may be used to detect when motion stops. Alternatively or additionally, a fixed or variable time delay may be added before starting the biological measurement. For example, the biological measurement may begin three seconds after motion is first detected, or the biological measurement may begin two seconds after motion is no longer detected.

[0032] During operation, performing biological measurements (512) may optionally include classifying pixels (514) and calculating a density based on the pixel classifications (516).

[0033] FIG. 6 illustrates an exemplary computing device 600 capable of implementing aspects of the present technology according to one or more implementations, including, for example, systems 200, 300, 400 (FIGS. 2-4) and method 500 (FIG. 5). Computing device 600 may be and / or be part of any computing device or server for producing the features and processes described above, including, but not limited to, a laptop computer, a smartphone, a tablet device, a wearable device such as goggles or eyeglasses, earphones or other audio devices, a case for an audio device, etc. Computing device 600 may include various types of computer-readable media and interfaces for various other types of computer-readable media. Computing device 600 includes a permanent storage device 602, a system memory 604 (and / or buffer), an input device interface 606, an output device interface 608, a bus 610, a ROM 612, one or more processing units 614, one or more network interfaces 616, and / or a subset or variation thereof.

[0034] The bus 610 collectively represents all system, peripheral, and chipset buses that communicatively connect the various internal devices of the computing device 600. In one or more implementations, the bus 610 communicatively connects one or more processing units 614 with the ROM 612, the system memory 604, and the permanent storage device 602. From these various memory units, the one or more processing units 614 retrieve instructions to execute and data to process in order to perform the processes of the present disclosure. The one or more processing units 614 may, in different implementations, be a single processor or a multi-core processor.

[0035] ROM 612 stores static data and instructions needed by one or more processing units 614 and other modules of computing device 600. Permanent storage device 602, on the other hand, may be a memory device that can be read and written. Permanent storage device 602 may be a non-volatile memory unit that stores instructions and data even when computing device 600 is off. In one or more implementations, mass storage devices (such as magnetic or optical disks and their corresponding disk drives) may be used as permanent storage device 602.

[0036] In one or more implementations, a removable storage device (such as a floppy disk, flash drive, and its corresponding disk drive) can be used as the permanent storage device 602. Like the permanent storage device 602, the system memory 604 can be a read / write memory device. However, unlike the permanent storage device 602, the system memory 604 can be a volatile read / write memory, such as a random access memory. The system memory 604 can store any instructions and data that the one or more processing units 614 may need during execution. In one or more implementations, the processes of the present disclosure are stored in the system memory 604, the permanent storage device 602, and / or the ROM 612. From these various memory units, the one or more processing units 614 retrieve instructions to execute and data to process to perform the processes of one or more implementations.

[0037] The bus 610 also connects to an input device interface 606 and an output device interface 608. The input device interface 606 enables a user to communicate information and select commands to the computing device 600. Input devices that can be used with the input device interface 606 may include, for example, an alphanumeric keyboard and a pointing device (also called a "cursor control device"). The output device interface 608 may enable, for example, the display of images generated by the computing device 600. Output devices that can be used with the output device interface 608 may include, for example, printers and display devices, such as liquid crystal displays (LCDs), light emitting diode (LED) displays, organic light emitting diode (OLED) displays, flexible displays, flat panel displays, solid-state displays, projectors, or any other device for outputting information.

[0038] One or more embodiments may include a device that functions as both an input device and an output device, such as a touchscreen. In these implementations, the feedback provided to the user may be any form of sensory feedback, such as visual feedback, auditory feedback, or haptic feedback, while input from the user may be received in any form, including acoustic, speech, or tactile input.

[0039] 6, bus 610 also couples computing device 600 to one or more networks and / or one or more network nodes via one or more network interfaces 616. In this manner, computing device 600 may be part of a network of computers, such as a local area network ("LAN"), a wide area network ("WAN"), an intranet, or the Internet. Any or all components of computing device 600 may be used in conjunction with the present disclosure.

[0040] Implementations within the scope of this disclosure may be realized in part or in whole using a tangible computer-readable storage medium (or multiple tangible computer-readable storage media of one or more types) encoding one or more instructions. The tangible computer-readable storage medium may also be non-transitory in nature.

[0041] A computer-readable storage medium may be any storage medium that can be read, written, or otherwise accessed by a general-purpose or special-purpose computing device, including any processing electronics and / or processing circuitry capable of executing instructions. For example, but not limited to, a computer-readable medium may include any volatile semiconductor memory, such as RAM, DRAM, SRAM, T-RAM, Z-RAM, and TTRAM. A computer-readable medium may also include any non-volatile semiconductor memory, such as ROM, PROM, EPROM, EEPROM, NVRAM, flash, nvSRAM, FeRAM, FeTRAM, MRAM, PRAM, CBRAM, SONOS, RRAM, NRAM, Racetrack memory, FJG, and Millipede memory.

[0042] Additionally, the computer-readable storage medium may include any non-semiconductor memory, such as optical disk storage, magnetic disk storage, magnetic tape, other magnetic storage devices, or any other medium capable of storing one or more instructions. In one or more implementations, the tangible computer-readable storage medium may be directly coupled to a computing device, while in other implementations, the tangible computer-readable storage medium may be indirectly coupled to a computing device, for example, via one or more wired connections, one or more wireless connections, or any combination thereof.

[0043] The instructions may be directly executable or may be used to develop executable instructions. For example, the instructions may be implemented as executable or non-executable machine code, or as instructions in a high-level language that can be compiled to generate executable or non-executable machine code. Furthermore, the instructions may be implemented as or include data. Computer-executable instructions may also be organized in any format, including, for example, routines, subroutines, programs, data structures, objects, modules, applications, applets, functions, and the like. Those skilled in the art will recognize that details such as the number, structure, order, and arrangement of instructions may vary considerably without changing the underlying logic, function, processing, or output.

[0044] While the above description primarily refers to microprocessors or multi-core processors executing software, one or more implementations are performed by one or more integrated circuits, such as ASICs or FPGAs, which execute instructions stored on the circuitry itself.

[0045] Those skilled in the art will understand that the various illustrative blocks, modules, elements, components, methods, and algorithms described herein can be implemented as electronic hardware, computer software, or a combination of both. To illustrate this interchangeability of hardware and software, the various illustrative blocks, modules, elements, components, methods, and algorithms have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the particular application and design constraints imposed on the overall system. Those skilled in the art will be able to implement the described functionality in a variety of ways for each particular application. The various components and blocks can be arranged differently (e.g., arranged in a different order or divided in a different way), all without departing from the scope of the present technology.

[0046] It should be understood that any particular order or hierarchy of blocks in the disclosed processes is an illustration of an example approach. Based on design preferences, it should be understood that the particular order or hierarchy of blocks in the processes may be rearranged, or that all depicted blocks may be executed. Any of the blocks may be executed simultaneously. In one or more implementations, multitasking and parallel processing may be advantageous. Furthermore, it should be understood that the separation of various system components in the above-described implementations should not be understood as requiring such separation in all implementations, and that the described program components (e.g., computer program products) and systems may generally be integrated together in a single software product or packaged in multiple software products.

[0047] As used herein and in any claims of this application, the terms "base station," "receiver," "computer," "server," "processor," and "memory" all refer to electronic or other technological devices. These terms exclude people or groups of people. For purposes of this specification, the terms "display" or "displaying" mean displaying on an electronic device.

[0048] As used herein, the phrase "at least one of," preceding a list of items, with the term "and" or "or" separating any of the items, modifies the list as a whole, not each member (i.e., each item) of the list. The phrase "at least one" does not require the selection of at least one of each listed item. Rather, the phrase allows for a meaning including at least one of any one of the items, and / or at least one of any combination of the items, and / or at least one of each of the items. By way of example, the phrase "at least one of A, B, and C" or "at least one of A, B, or C" refers, respectively, to A only, B only, or C only, any combination of A, B, and C; and / or at least one of each of A, B, and C.

[0049] The predicates "configured to," "operably," and "programmed to" do not imply any particular tangible or intangible modification of the subject matter, but rather are intended to be used interchangeably. In one or more implementations, a processor configured to monitor and control operations or components can also mean that the processor is programmed to monitor and control operations or that the processor is operable to monitor and control operations. Similarly, a processor configured to execute code can be interpreted as a processor programmed to execute code or operable to execute code.

[0050] Phrases such as "one aspect," "that aspect," "another aspect," "some aspects," "one or more aspects," "one implementation aspect," "that implementation aspect," "another implementation aspect," "some implementation aspects," "one or more implementation aspects," "one embodiment," "that embodiment," "another embodiment," "some implementation aspects," "one or more implementation aspects," "configuration," "that configuration," "another configuration," "some configurations," "one or more configurations," the technology, the disclosure, the disclosure, and other variations thereof are used for convenience and do not imply that disclosure associated with such phrases is essential to the technology or that such disclosure applies to all configurations of the technology. Disclosure associated with such phrases may apply to all configurations or to one or more configurations. Disclosure associated with such phrases may provide one or more examples. Phrases such as "one aspect" or "some aspects" can refer to one or more aspects, and vice versa, and this applies equally to other aforementioned phrases.

[0051] The word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any embodiment described herein as "exemplary" or "example" is not necessarily to be construed as preferred or advantageous over other implementations. Furthermore, to the extent that terms like "include," "have," and the like are used in the specification or claims, such terms are intended to be inclusive in the same manner as the term "comprising," such as "comprise" when used as a transitional term in a claim.

[0052] All structural and functional equivalents to the elements of the various aspects described throughout this disclosure that are known or later become known to those skilled in the art are expressly incorporated herein by reference and are intended to be encompassed by the claims. Furthermore, nothing disclosed herein is intended to be dedicated to the public, regardless of whether such disclosure is expressly recited in the claims. No claim element is to be construed under the provisions of U.S.C. Section 112(f) unless the element is expressly recited using the phrase "means for," or, in the case of a method claim, unless the element is recited using the phrase "step for."

[0053] The above description is provided to enable those skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles herein may be applied to other aspects. Accordingly, the claims are not limited to the aspects set forth herein but are to be accorded the full scope consistent with the language of the claims, and references to elements in the singular do not mean "one and only one," unless so expressly stated, but rather "one or more." Unless expressly stated otherwise, the term "some" refers to one or more. Masculine pronouns (e.g., his) include feminine and neuter forms (e.g., her and its), and vice versa. Headings and subheadings, where used, are used for convenience only and are not intended to limit the disclosure.

Claims

1. 1. An image processing method, comprising: determining a mutual information metric between at least one pair of images from a sequence of images of the biological sample; detecting movement of the biological specimen based on the mutual information metric; after the motion is detected, performing a measurement of the biological specimen based on a test image of the image sequence.

2. The image processing method of claim 1 , further comprising forgoing the performing of the measurement of the biological sample if the movement is not detected.

3. The image processing method of claim 1 , wherein said determining said mutual information metric comprises estimating a measure of statistical independence between co-located pixel values ​​in said image pair.

4. The image processing method of claim 3 , wherein said estimating said measure of statistical independence comprises determining a joint histogram of said co-located pixel values.

5. The image processing method of claim 1 , wherein the motion is detected when the mutual information metric exceeds a mutual information threshold level.

6. The image processing method of claim 1 , wherein the motion is detected based on multiple metrics of mutual information, each of the multiple metrics being between a different pair of images from the image sequence.

7. 2. The image processing method of claim 1, wherein said performing said measurement of said biological sample is delayed after said motion is detected until said motion is no longer detected.

8. said performing said measurement of said biological sample comprises: The image processing method of claim 1 , comprising processing the test image with a machine learning model to generate the measurement of the biological sample.

9. said performing said measurement of said biological sample comprises: analyzing the test image from the image sequence to generate a plurality of feature images; deriving a likelihood image for each of a plurality of object types from the feature image; 2. The image processing method of claim 1, further comprising: combining the likelihood image into a classification image indicating which of the plurality of object types is detected at each pixel in the classification image.

10. The image processing method of claim 9 , further comprising calculating a density metric for an object type based on the percentage of pixels in the classification image that exhibit said object type.

11. 10. The image processing method of claim 9, wherein the plurality of feature images includes an average image having pixel values ​​each based on an average of neighboring pixels of the test image, and a standard deviation image having pixel values ​​each based on a standard deviation of neighboring pixels of the test image.

12. The image processing method of claim 1 , wherein the sequence of images of the biological specimen is a sequence of transmitted light images captured by a camera with illumination behind the biological specimen.

13. An image processing device comprising a controller configured to perform the method of any one of claims 1 to 12.

14. A non-transitory computer readable medium storing instructions that, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 12.

Images