Machine-readable markers and identification methods

Abstract

Broadly speaking, embodiments of the present technology provide optically readable or machine readable fiducial markers for use in nanoscale applications and methods for automated detection of the markers. Advantageously, the fiducial markers of the present technology have characteristics that make them resistant to the processing steps of electron beam lithography. Furthermore, the fiducial markers are robust and have high information density when reproduced at the nanoscale.

Description

[Technical field]

[0001] Generally, the present technology relates to machine-readable fiducial markers and methods for reading or identifying the markers. In particular, the present technology provides optically readable or machine-readable fiducial markers for use in, for example, nanoscale applications, and methods for automated detection of the markers. [Background technology]

[0002] To reduce power consumption or to achieve a globally sustainable power consumption level, it is desirable to create efficient optoelectronic devices. To make optoelectronics more efficient, atomically thin nanomaterials and specific microstructures, e.g., single-wall, grain boundary, or nanowires / nanotubes, are required. In this case, nanowires can be used, for example, to form highly efficient photovoltaic devices.

[0003] Generally, there are two approaches to constructing a nanofeature of interest: top-down and bottom-up. In the top-down approach, the nanofeature of interest is formed by reducing a larger piece of material, for example by lithography. In the bottom-up approach, the nanofeature of interest is synthesized by combining constituent atoms. These bottom-up processes can include growing the nanofeature of interest.

[0004] When nanofeatures of interest are grown on a substrate, they must be located before they can be used to build more complex structures, such as, for example, optoelectronic devices. However, currently, the identification of these nanostructures relies on human input and is therefore both time consuming and expensive. If the process of identifying nanofeatures of interest could be automated to enable the use of these nanofeatures for the manufacture of, for example, optoelectronics, it would take less time to manufacture efficient optoelectronics and therefore become commercially viable.

[0005] Fiducial markers are useful tools to aid in the automation of identification of nanofeatures of interest, and thus provide a significant improvement over current processes in nanofabrication, such as lithography, development, lift-off metallization, etc. Such fiducial markers can be used as alignment markers that can be automatically identified using computer vision methods. In this case, these alignment markers can be used to locate features of interest on a sample or substrate and achieve alignment of successive layers of lithographically defined nanostructures, such as electrodes or antennas.

[0006] However, to be useful, any alignment / fiducial marker must remain as unaltered as possible during the nanofabrication process (e.g., lithography, etc.). In addition to being resistant to alteration by manufacturing process steps, the fiducial marker must be sufficiently amenable to detection by computer vision methods. Current fiducial markers do not meet these criteria and / or are not suitable for use in the creation of nanoscale features. Summary of the Invention [Problem to be solved by the invention]

[0007] Accordingly, Applicants have come to recognize a need for improved fiducial markers and methods for reading or identifying the fiducial markers. [Means for solving the problem]

[0008] In a first approach of the present technology, a machine-readable fiducial marker is provided. The marker comprises a spine having a minor axis and a major axis, the orientation of which defines the orientation of the marker on a substrate. The spine defines the shape of the marker and at least two regions of the marker having a perimeter-to-area ratio within a predetermined threshold range. Additionally, the marker comprises at least one set of symbols disposed in one or more of the at least two regions, the at least one set of symbols encoding coordinate information of a location relative to the centroid of the spine on the substrate.

[0009] The fiducial markers may be suitable for use in nanoscale applications, although it will be understood that this is a non-limiting example of use, and in fact the fiducial markers may be scaled as needed for various applications.

[0010] These symbols may encode coordinate information of the centroid of the spine (in this case the position relative to the centroid is the centroid position), or coordinate information of another position on the substrate (in this case the position relative to the centroid is another position on the substrate away from the centroid).

[0011] The shape of the marker may be any two-dimensional or planar shape, such as a circle, a shape having at least one curved edge or side, and a polygon.

[0012] The perimeter to area ratio defined by the spine may indicate the circularity of the shape.

[0013] At least one set of symbols may encode coordinate information of the centroid of the fiducial marker's spine, or may encode coordinate information of another location / position relative to the centroid.

[0014] Thus, in some cases, the at least one set of symbols may comprise a first set of symbols that encode an x-coordinate of the centroid of the spine and a second set of symbols that encode a y-coordinate of the centroid of the spine.

[0015] In these cases, the at least one set of symbols may further comprise a third set of symbols encoding a checksum with respect to the x coordinate and a fourth set of symbols encoding a checksum with respect to the y coordinate.

[0016] In other cases, the at least one set of symbols may comprise a first set of symbols that encode a distance from the centroid of the spine and a second set of symbols that encode an angle from a reference direction defined by the spine, which may be defined by the long axis of the spine, i.e., the symbols may encode polar coordinates of the centroid of the spine instead of Cartesian coordinates.

[0017] In these cases, the at least one set of symbols may comprise a third set of symbols that encodes a checksum over distance and a fourth set of symbols that encodes a checksum over angle.

[0018] In other cases, the at least one set of symbols may comprise a first set of symbols encoding a first coordinate of an offset position relative to the centroid of the spine and a second set of symbols encoding a second coordinate of the offset position relative to the centroid of the spine.

[0019] In such cases, the first and second coordinates may be Cartesian or polar coordinates. The at least one set of symbols may comprise a third set of symbols encoding a checksum for the first coordinate and a fourth set of symbols encoding a checksum for the second coordinate. As will be described in more detail with reference to the drawings, the first and second coordinates may form instructions for a machine in a calibration process or a manufacturing process.

[0020] The spine may comprise at least two arms. The spacing between each symbol in the at least one set of symbols may be the same as the width of each arm of the spine.

[0021] The spine may define a hexagonal marker, a square marker, or a rectangular marker, it being understood that these are non-limiting examples of the shapes that the marker may take.

[0022] At least one set of symbols encoding coordinate information may be arranged in a predetermined pattern that indicates a fixed position for each symbol. For example, the predetermined pattern that indicates a fixed position for each symbol may consist of 4, 9, or 16 positions. It will be understood that these are non-limiting examples of the number of positions. The number of positions may depend on the overall size and / or overall shape of the marker. For example, a larger marker may have a larger area within which a larger number of fixed positions / symbols may be provided. However, it is not necessary that a larger marker has a larger number of fixed positions / symbols.

[0023] In one particular example of a fiducial marker configuration, the spine may be a cross-shaped spine having four arms and defining four regions of the marker. In this example, the at least one set of symbols may comprise a first set of symbols provided in a first region of the four regions, the first set of symbols being arranged in a predetermined pattern indicative of a fixed position of each symbol and forming a binary representation of a first coordinate of a position relative to the centroid of the spine on the substrate, a second set of symbols provided in a second region of the four regions, the second set of symbols being arranged in a predetermined pattern indicative of a fixed position of each symbol and forming a binary representation of a second coordinate of a position relative to the centroid of the fiducial marker on the substrate, a third set of symbols provided in a third region of the four regions, the third set of symbols defining a checksum for the first coordinate, and a fourth set of symbols provided in a fourth region of the four regions, the fourth set of symbols defining a checksum for the second coordinate.

[0024] In a second approach of the present technology, an apparatus is provided. The apparatus includes a substrate and at least one machine-readable fiducial marker. The at least one machine-readable fiducial marker includes a spine having a minor axis and a major axis, the orientation of the major axis and the minor axis defining the orientation of the marker on the substrate. The spine defines a shape of the marker having a perimeter-to-area ratio within a predetermined threshold range and at least two regions of the marker. Furthermore, the at least one machine-readable fiducial marker includes at least one set of symbols disposed in one or more of the at least two regions. The at least one set of symbols encodes coordinate information of a location relative to the centroid of the spine on the substrate.

[0025] The features described above with reference to the first approach apply equally to the machine-readable fiducial markers of the second approach and therefore will not be repeated for the sake of clarity.

[0026] The at least one machine-readable fiducial marker of the device may be an array of fiducial markers. The array of fiducial markers may comprise fiducial markers of at least two different sizes.

[0027] The substrate of the apparatus may be any one of a patternable material, a semiconductor device, a semiconductor wafer, a metal layer, a dielectric layer, a transparent substrate, and a microscope cover slip.

[0028] In some cases, the at least one machine-readable fiducial marker may be formed by patterning the substrate with a contrast material, It will be understood that the fiducial marker(s) may be formed using any suitable manufacturing process or technique.

[0029] In a third approach of the present technology, a method for identifying machine-readable fiducial markers of the type described herein is provided, the method comprising acquiring an image of a substrate on which the machine-readable fiducial markers are formed, acquiring information about properties of the fiducial markers formed on the substrate, and using the acquired information to identify the fiducial markers in the image.

[0030] Obtaining information about a characteristic of the fiducial marker may include obtaining a perimeter-to-area ratio of a shape defined by a spine of the fiducial marker.

[0031] The method may further include the steps of identifying a potential fiducial marker in the image, determining a centroid of the potential fiducial marker, determining a major axis and a minor axis of a spine of the potential fiducial marker using the centroid, calculating a perimeter-to-area ratio of a potential shape defined by the spine using the determined major axis and minor axis, determining whether the calculated perimeter-to-area ratio matches a ratio in the acquired information, and identifying the potential fiducial marker as a fiducial marker if it is determined that the calculated perimeter-to-area ratio matches the ratio in the acquired information.

[0032] Additionally, the method may include determining an orientation of the fiducial marker using the major and minor axes.

[0033] The method may further include the steps of determining fixed positions of at least one set of symbols encoding coordinate information using the acquired information, identifying whether any symbols are present at the fixed positions based on the fixed positions of the symbols, and reading first and second coordinates of a position relative to the centroid of the fiducial marker based on the identification of any symbols.

[0034] Obtaining information about a characteristic of the fiducial marker may include obtaining a template of a spine of the fiducial marker.

[0035] In this case, the method may further include the steps of identifying a potential reference marker in the image, rotating and scaling the template to match the orientation and size of the potential reference marker, calculating a correlation value between the identified potential reference marker in the image and the template, and identifying the potential reference marker as a reference marker if the correlation value indicates that the potential reference marker closely matches the template.

[0036] The method may further include using the template to identify whether any symbol is present in a fixed position, and reading first and second coordinates of a position relative to the centroid of the reference marker based on the identification of the any symbol.

[0037] The method may further include the steps of using the acquired information to identify any symbols encoding a checksum for the first coordinate and a checksum for the second coordinate and reading both of these checksums, calculating the checksums for the first coordinate and the second coordinate, and checking whether the calculated checksums match the read checksums, where if the calculated checksum matches the read checksum the potential fiducial marker has been correctly identified and read.

[0038] In a fourth approach of the present technology, a method is provided for controlling a machine during a machine calibration or manufacturing process using an array of machine-readable fiducial markers of the type provided herein on a substrate, the method comprising capturing at least one image of the array of machine-readable fiducial markers, identifying at least one machine-readable fiducial marker in the array using a method provided herein, and controlling the machine using the identified at least one fiducial marker.

[0039] The method may be for calibrating a translation stage of an imaging machine, in which case the method may include reading an identified fiducial marker to determine first and second coordinates of a location relative to a centroid of the fiducial marker, moving the translation stage a predetermined distance, determining a distance traveled by the translation stage relative to the identified fiducial marker using an array of machine readable fiducial markers, calculating an actual distance traveled by the translation stage, and calibrating the translation stage using the actual distance.

[0040] The method may be for calibrating the magnification of an imaging machine, where the method includes using an array of machine-readable fiducial markers to determine spacing between the fiducial markers in pixels, and calculating the magnification of the imaging machine using the known distances between the fiducial markers in the array.

[0041] The method may be for calibrating a focus of an imaging machine, in which case the method may include convolving the captured image with a template of a spine of a fiducial marker, identifying a peak in the convolution indicative of a presence of the fiducial marker, and controlling the imaging machine to adjust the focus to increase the peak in the convolution.

[0042] In one related approach to the present technology, a non-transitory data carrier is provided that carries processor control code for implementing any of the methods, processes and techniques described herein.

[0043] As will be appreciated by one skilled in the art, the present technology may be embodied as a system, method, or computer program product, and thus may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0044] Furthermore, the present technology may take the form of a computer program product embodied in a computer readable medium having computer readable program code embodied therein. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing.

[0045] Computer program code for carrying out operations of the present technology may be written in any combination of one or more programming languages, including object-oriented and conventional procedural programming languages. Code components may be embodied as procedures or methods, etc., and may include subcomponents that may take the form of instructions or sequences of instructions at any level of abstraction, from direct machine instructions of a native instruction set to higher level compiled or interpreted language constructs.

[0046] Furthermore, embodiments of the present technology provide a non-transitory data carrier carrying code which, when implemented in a processor, causes the processor to perform any of the methods described herein.

[0047] Further, these techniques provide processor control code for implementing the above-mentioned methods, such as in a general-purpose computer system or a digital signal processor (DSP). Further, the present techniques provide a carrier carrying processor control code for implementing, when executed, any of the above-mentioned methods, particularly on a non-transitory data carrier. This code may be provided on a carrier such as a disk, a microprocessor, a CD-ROM, a DVD-ROM, a programmed memory, such as a non-volatile memory (e.g. flash memory) or a read-only memory (firmware), or on a data carrier, such as an optical or electrical signal carrier. Code (and / or data) for implementing embodiments of the techniques described herein may include source, object, or executable code in a conventional programming language (interpreted or compiled), such as C, for example, or assembly code, code for configuring or controlling an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array), or code for a hardware description language, such as Verilog (RTM) or VHDL (Very High Speed ​​Integrated Circuit Hardware Description Language). As will be appreciated by those skilled in the art, such code and / or data may be distributed among multiple coupled components in communication with each other. The technology may include a controller that includes a microprocessor, working memory, and program memory coupled to one or more of the components of the system.

[0048] Moreover, it will also be apparent to those skilled in the art that all or part of the logical methods according to embodiments of the present technology may be suitably embodied in logic devices comprising logic elements for performing the steps of the above-described methods, and that such logic elements may comprise components such as logic gates in, for example, programmable logic arrays or application specific integrated circuits, etc. Such logic configurations may further be embodied in enabling elements for establishing logic structures temporarily or permanently within such arrays or circuits using, for example, virtual hardware description languages, which may be stored and transmitted using fixed or transmittable carrier media.

[0049] In one embodiment, the techniques may be implemented using multiple processors or control circuits. The techniques may be configured to run on or be integrated into the device's operating system.

[0050] In one embodiment, the technology may be realized in the form of a data carrier carrying functional data, said functional data comprising functional computer data structures for enabling said computer system to carry out all the steps of the above-mentioned method when loaded into and operated by a computer system or network.

[0051] Implementations of the present technology will now be described, by way of example only, with reference to the accompanying drawings, in which: [Brief description of the drawings]

[0052] [Figure 1] FIG. 1 is a schematic diagram illustrating one exemplary method for identifying nanofeatures of interest. [Figure 2A] 1A-1C show an example of a fiducial marker design and its appearance when formed using nanofabrication techniques. [Figure 2B] 1A-1C show an example of a fiducial marker design and its appearance when formed using nanofabrication techniques. [Figure 2C] 1A-1C show an example of a fiducial marker design and its appearance when formed using nanofabrication techniques. [Figure 3A] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 3B] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 3C] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 3D] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 3E] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 3F] 1A-1C show examples of machine-readable fiducial markers according to the present technology. [Figure 4A] FIG. 1 illustrates an example of a fiducial marker according to the present technology. [Figure 4B] 1 is a schematic diagram illustrating how fiducial markers may encode Cartesian coordinate information. [Figure 4C] FIG. 1 is a schematic diagram illustrating how a fiducial marker may encode polar coordinate information. [Figure 4D] FIG. 13 is a schematic diagram illustrating how a fiducial marker may encode coordinate information regarding a position offset from a marker. [Figure 4E] FIG. 1 is a schematic diagram showing how a fiducial marker may encode instructions for a machine. [Diagram 5] 1A-1C show an example design and appearance of a fiducial marker when formed using nanofabrication techniques. [Figure 6A] FIG. 1 illustrates one way in which fiducial markers can be fabricated and imaged. [Figure 6B] FIG. 1 illustrates one way in which fiducial markers can be fabricated and imaged. [Figure 6C] FIG. 1 illustrates one way in which fiducial markers can be fabricated and imaged. [Figure 6D] FIG. 1 illustrates one way in which fiducial markers can be fabricated and imaged. [Figure 7] FIG. 1 shows a flow diagram of example steps for identifying machine-readable fiducial markers of the present technology. [Figure 8] FIG. 13 shows a flow diagram of example steps for identifying a machine-readable fiducial marker when the perimeter-to-area ratio of the marker is known. [Figure 9] FIG. 1 illustrates some pre-processing steps that may be performed before identification of machine-readable fiducial markers and a method that uses the perimeter-to-area ratio of the markers. [Figure 10] FIG. 13 shows a flow diagram of example steps for identifying a machine-readable fiducial marker when a template of the marker is known. [Figure 11] FIG. 1 illustrates how machine-readable fiducial markers are identified using a template. [Figure 12] 1 is an image of an array with two different sized fiducial markers. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0053] Broadly speaking, embodiments of the present technology provide optically or machine-readable fiducial markers and methods for automated detection of the markers. Advantageously, the fiducial markers of the present technology have characteristics that allow them to withstand the processing steps of electron beam lithography. Furthermore, the fiducial markers are robust and have high information density when reproduced at the nanoscale. The fiducial markers may be suitable for use in nanoscale applications, although it will be understood that this is a non-limiting example of use. Indeed, the fiducial markers may be scaled up or down as needed for various applications.

[0054] FIG. 1 is a schematic diagram showing one typical method for identifying nanofeatures of interest. This typical method is performed by a human. These nanofeatures may be, for example, nanowires on a silicon wafer. The nanofeatures can be prepared on a substrate during a sample preparation stage. The sample can then be investigated using imaging and / or microscopy techniques to, for example, identify the nanofeatures. Once identified, these nanofeatures can be used to build electronic or optoelectronic devices. However, to manufacture these devices, the exact location and orientation of the nanofeatures must be known so that further processing techniques can be applied to the exact location on the substrate.

[0055] Fiducial markers are man-made patterns, symbols, or images placed in an environment to be used for reference or image recognition. They are most common in augmented reality applications and are used for a variety of computer vision tasks including object tracking, image calibration, and localization. Once fiducial markers are recognized by an imaging system, they can be used to determine the location and orientation of the object to which they are attached relative to the imaging system. The performance of fiducial markers can be evaluated by: 1) the false positive rate, where a marker is detected when it is not present; 2) the inter-marker confusion rate, where one marker is mistaken for another; 3) the false negative rate, where a marker is present but not detected; 4) the minimum marker size, which is the pixel size required for accurate detection of the marker; and 5) the sensitivity to imaging conditions. There are many different fiducial markers, and their design is highly application dependent. Some fiducial marker patterns are specifically designed to increase information density, while others are developed to minimize the false negative rate and reduce the minimum marker size. Fiducial markers can have significant advantages in device manufacturing processes for electronic applications due to their repeatability, high throughput, and high level of accuracy, and can aid in the automation of many activities that currently require manual intervention.

[0056] The shape and geometry of fiducial markers are crucial as design features such as circles, lines, cones, etc. are used to characterize and detect reference points between markers. Visual fiducial markers generally contain less information than 2D barcode markers, but still provide information about their position relative to the imaging system and other markers.

[0057] 2A-2C show several known fiducial marker designs and their appearance when formed using nanofabrication techniques.

[0058] Figure 2A shows, on the left side, one of the first fiducial marker designs, which comprises symmetrical concentric circles with alternating black and white circles.

[0059] Figure 2B shows an example of a fiducial marker, known as AprilTag, on the left. This fiducial marker uses a binary internal system for signaling. However, drawbacks of this type of marker include a false positive detection rate that increases as the size of the marker decreases, and high sensitivity to imaging conditions.

[0060] Figure 2C shows, on the left, an example of a fiducial marker that aims to address these shortcomings of AprilTag. Known as RuneTag, this marker uses a pattern of circular dots of various sizes, which results in a significantly improved detection rate.

[0061] The images on the right in Figures 2A-C show what each of these known fiducial markers look like when fabricated on a silicon substrate using sputtered tungsten. Fiducial markers of different sizes (2 μm and 20 μm) were fabricated and then imaging techniques were used to determine how well these markers were formed. These images on the right were acquired using a scanning electron microscope (SEM).

[0062] From these SEM images, it is clear that not all known fiducial markers are suitable for nanofabrication techniques, because the shape of these fiducial markers is easily destroyed or altered during processing. For example, the shape of the Concentric Contrasting Circles (CCC) fiducial markers in FIG. 2A is significantly altered during the nanofabrication process. Furthermore, bridging occurs between features as a result of proximity effects or densification of exposed features. Similarly, the SEM image in FIG. 2B shows damaged edges. Since these edges of the markers are used for identification, such damaged markers are not suitable for use in small-scale applications, e.g., nanofabrication processes. The SEM image in FIG. 2C shows that RuneTag is not suitable for nanofabrication processes.

[0063] These experiments with existing fiducial markers reveal a set of requirements that a marker suitable for nanoscale applications must fulfill. General requirements for fiducial markers suitable for nanoscale applications include robustness, high recognition reliability, and insensitivity to lighting or imaging conditions. In addition, there are requirements specific to nanofabrication processes, such as lithography, development, and lift-off metallization. Fiducial markers in lithography are used as alignment markers. Alignment markers are used to locate features of interest on a sample, such as nanostructured materials, and to achieve alignment of successive layers of lithographically defined nanostructures, such as electrodes or antennas. In nanodevice manufacturing, these features are on the nanometer scale, so the alignment markers themselves need to be of a similar size and maintain high resolution throughout the processing steps to achieve high recognition reliability.

[0064] Many factors can affect the accuracy of patterns in nanofabrication processes. The accuracy of the pattern size in electron beam lithography is determined by the size of the electron beam, the properties of the resist, and the proximity effect caused by indirect exposure due to scattered electrons in the resist and substrate. Cross-type is a common alignment marker used in lithography, but its proximity effect can cause the exposed area to be larger than expected as well as oversaturating the camera, which can result in missing features of the pattern, changing the pattern design, and making the alignment marker position uncertain. Another resolution-limiting process in nanofabrication is lift-off. This lift-off is used to transfer the pattern by evaporating metal on the resist and then dissolving the resist to leave only the deposited metal that was in contact with the substrate. To achieve a clean lift-off, an undercut profile should be formed by avoiding deposition on the sidewalls by choosing the appropriate resist. This is a common challenge when ridges form on the sidewalls due to stochastic fluctuations and noise effects during the electron beam lithography process. Both of these phenomena become more pronounced as the feature dimensions shrink.

[0065] To minimize overexposure and ensure that the fiducial marker retains its shape during lift-off, certain geometric requirements for the fiducial marker design include avoiding closed open spaces in the pattern and having all features larger than the minimum achievable resolution of the system. Additionally, the fiducial marker should have easily identifiable features for use as reference points to minimize alignment uncertainty. Additionally, from a computer vision perspective, the fiducial marker's shape should be easily identifiable from the surrounding environment, provide good material contrast, and have sufficient information density.

[0066] 3A-3F show examples of machine-readable fiducial markers, referred to as "LithoTags," in accordance with the present technology. The present technology provides machine-readable fiducial markers comprising a marker shape having a perimeter-to-area ratio within a predetermined threshold range and a spine defining at least two regions of the marker, and at least one set of symbols disposed within one or more of the at least two regions, the at least one set of symbols encoding coordinate information of a location relative to the centroid of the spine on a substrate.

[0067] For typical lithographic processing techniques, the LithoTag design depicts a locally patterned contrast material (CM) (e.g., a physical vapor deposited metal) on a uniform global background, such as a wafer substrate. In this case, the CM is selected to maximize contrast in a given imaging technique, including optical microscopy (OM), scanning electron microscopy (SEM), or atomic force microscopy (AFM). The width of the spine arms may be the same as the spacing between symbols in the x and y directions.

[0068] In the examples shown in Figures 3A-3C, the spine of the fiducial marker defines a hexagonal shape and four regions located within the shape / fiducial marker. In the case of a hexagonal shape, the spine has four arms and has an aspect ratio and circularity that aids in recognition. The spine may provide information regarding the orientation of the tag with its longest arm pointing / indicating the top of the marker. In these examples, two regions on the left side of the marker provide information regarding the x coordinate of the marker on the substrate, and two regions on the right side of the marker provide information regarding the y coordinate of the marker on the substrate. However, it will be appreciated that one region may be used to provide information regarding the x coordinate and another region may be used to provide information regarding the y coordinate. Thus, as previously discussed, the fiducial marker may have more than one region defined by a spine.

[0069] It can be seen that the number of symbols (circles / dots) provided in these regions increases from Figure 3A to Figure 3C. In Figure 3A, the top left and top right regions have space for 4 symbols each, in Figure 3B, these regions have space for 9 symbols each, and in Figure 3C, these regions may have space for 16 symbols each. Each symbol can be considered as 1 bit of information. The symbols in each region may be arranged according to a predefined pattern that indicates the fixed position of each symbol. Thus, the presence or absence of a symbol at each fixed position in each region may encode coordinate information. The more symbols there are, the more coordinates can be encoded. In the case of the marker in Figure 3C with 16 bits in each region, this marker can encode up to (2 16 ) 2 possible coordinate combinations can be encoded, i.e. ensuring high information density with a maximum of 4294967296 possible coordinate combinations. Thus, the fiducial markers of the present technology can have high information density.

[0070] In the example shown in Figures 3D-3F, the fiducial marker spine defines a rectangular shape and four regions located within the shape / fiducial marker. Again, it can be seen that the number of symbols provided within these regions increases from Figure 3D-3F. The rectangular shape is designed to accommodate lower feature resolution requirements, such as for optical lithography applications.

[0071] It will be understood that the designs shown in Figures 3A-3F are merely examples of fiducial marker designs of the present technology and are not limiting.

[0072] FIG. 4A shows a fiducial marker 100 of an example of the present technology in more detail. This fiducial marker 100 is identical to the fiducial marker shown in FIG. 3B. The fiducial marker 100 comprises a spine 110 that defines the shape of the marker 100. In this case, the shape defined by the spine 100 is a hexagon. However, the shape of the marker may be any two-dimensional shape or flat figure, such as a circle, a shape with at least one curved edge or side, and a polygon. Thus, the shapes shown herein are not limiting. The shape may have a perimeter-to-area ratio within a predetermined threshold range. In some cases, the perimeter-to-area ratio defined by the spine may indicate the circularity of the shape.

[0073] Furthermore, the spine 110 defines at least two regions of the marker 100. In this example, the spine 110 defines four regions 102-108.

[0074] When using a function to draw a LithoTag, centroid coordinates are provided, indicating the location of the tag's centroid on the x and y axes, and the tag type (4, 9, or 16), scale factor, and tag orientation must be specified.

[0075] Thus, the marker 100 comprises at least one set of symbols provided in one or more of the at least two regions of the marker 100, the at least one set of symbols encoding coordinate information of the centroid of the spine 110 on the substrate. In this case, there are four regions 102-108, and a set of symbols may be provided in each of the four regions.

[0076] The marker 100 may comprise a first set of symbols encoding the x-coordinate of the centroid of the spine 110, which may be located within region 102 of the marker 100, and a second set of symbols encoding the y-coordinate of the centroid of the spine 110, which may be located within region 104 of the marker 100.

[0077] Another important parameter to aid in the lithographic process is the fill factor, which is the percentage of marker patterned and deposited relative to the substrate material. Additionally, since the arms of the spine 110 use fillets, it is also possible to specify the number of arc points, which is the number of segments used to define the arc, and the number of circle points, which is the number of segments used to define the circle. Another design aspect of the marker is the number of inner arcs, which represents the number of points that define the inner fillet of the arm. The number of inner arcs is not required for recognition, but has been found to optimize lift-off. Not including this parameter results in fewer rows in the final object, which has the advantage of reducing the LithoTag generation time and reducing the size of the output file.

[0078] Checksums are a common method to check for information errors in a system. In the case of LithoTag, checksums are used for both the x and y coordinates, allowing to check if the tag was read correctly without errors. The checksum can be calculated from the coordinates using a cyclic redundancy check. By incorporating a checksum in the marker, it is possible to check for writing or reading errors. This gives information about compatibility with nanofabrication processes, including lift-off.

[0079] Thus, the marker 100 may further comprise a third set of symbols encoding a checksum for the x coordinate, which may be provided in region 106 of the marker 100, and a fourth set of symbols encoding a checksum for the y coordinate, which may be provided in region 108 of the marker 100.

[0080] As mentioned above, the LithoTag is designed to incorporate resistance to oversaturation that can result from imaging or overexposure / overdevelopment during lithography. The hexagonal design shown in Figure 4A minimizes the number of empty spaces on the substrate material, ensuring the highest packing density for compactness, and therefore minimizes proximity effects that can increase the size of the exposed pattern.

[0081] In comparison, the AprilTag marker (Figure 2B) contains a large proportion of closed open space, making it unsuitable for lithography applications. Similarly, in the RuneTag marker (Figure 2C), where the size and shape of the dots are critical for detection, the pattern can easily be altered during processing due to changes in process parameters, again unfavorable for lithography. In contrast, LithoTag is designed to overcome these issues due to its robustness against rough feature edges.

[0082] It is also important to consider the minimum resolution of the lithography system; that is, the smallest feature of the LithoTag marker must be larger than the minimum resolution. One advantage of the LithoTag design is that it can be easily scaled down to the nanometer scale and still maintain detection accuracy. Since the circle diameter is the size limiting factor, the smallest achievable circle size for the LithoTag is nearly the same as the minimum resolution, so for e-beam lithography of the marker in FIG. 3A with a spine width U of 10 nm, the overall minimum size of the LithoTag is less than 100 nm. The same minimum circle size for the RuneTag means that the overall marker is nearly twice the size of the LithoTag.

[0083] Previous fiducial markers, such as AprilTag (Figure 2B), have been designed to maintain detection accuracy against changes in projection angle, as they use sharp line edges and corners for detection. The LithoTag system does not use sharp edges, which are more likely to be damaged during processing. Instead, fill volume ratios are used for detection, which are less robust against changes in field projection angle, and therefore the markers can only be detected reliably when viewed from directly above. However, for nanofabrication purposes, features such as oversaturation resistance or minimum resolution are favorable tradeoffs against changes in projection angle field of view.

[0084] Thus, as shown in Figures 3A-4A, the fiducial marker spine 110 may comprise at least two arms. The spacing between each symbol in at least one set of symbols may be the same as the width U of each arm of the spine (as shown in Figure 4A). In some cases, the spine 110 may be a cross-shaped spine.

[0085] The spine 110 may define a hexagonal, square, or rectangular marker, it being understood that these shapes are non-limiting examples of the shapes the markers may take.

[0086] The spine 110 may have a short axis and a long axis, the orientation of which may define the orientation of a marker on a substrate. For example, as described above, the longest arm of the spine may define the apex of the marker and thus may be used to determine the orientation of the marker.

[0087] At least one set of symbols encoding coordinate information may be arranged in a predetermined pattern that indicates a fixed position for each symbol. For example, the predetermined pattern that indicates a fixed position for each symbol may consist of 4, 9, or 16 positions. In FIG. 4A, the predetermined pattern is shown by circles in each region 102-108. The circles are numbered in each region, and the numbers may indicate how to read the symbols in each region of the marker. It will be understood that these are non-limiting examples of the number of positions. The number of positions may depend on the overall size and / or shape of the marker. For example, a larger marker may have a larger region in which a larger number of fixed positions / symbols may be provided. However, it is not necessary that a larger marker has a larger number of fixed positions / symbols.

[0088] In one particular example of the fiducial marker configuration, the spine may be a cross-shaped spine having four arms and defining four regions of the marker. In this example, the at least one set of symbols may comprise a first set of symbols provided in a first region of the four regions, the first set of symbols being arranged in a predetermined pattern indicative of the location of each symbol and forming a binary representation of the x-coordinate of the centroid of the spine on the substrate, a second set of symbols provided in a second region of the four regions, the second set of symbols being arranged in a predetermined pattern indicative of the location of each symbol and forming a binary representation of the y-coordinate of the centroid of the fiducial marker on the substrate, a third set of symbols provided in a third region of the four regions, the third set of symbols defining a checksum for the x-coordinate of the centroid of the fiducial marker, and a fourth set of symbols provided in a fourth region of the four regions, the fourth set of symbols defining a checksum for the y-coordinate of the centroid of the fiducial marker. This example is illustrated in Figures 3B and 4A.

[0089] FIG. 4B is a schematic diagram showing how fiducial markers encode Cartesian coordinate information. As shown, a marker grid (right) showing the layout of markers (represented by stars) on a substrate may be patterned with regular intervals between each marker in the x and y directions. Each fiducial marker (left) then encodes integer values ​​for x and y. In such cases, the at least one set of symbols may comprise a first set of symbols that encodes the x coordinate of the centroid of the spine and a second set of symbols that encodes the y coordinate of the centroid of the spine. In these cases, the at least one set of symbols may further comprise a third set of symbols that encodes a checksum for the x coordinate and a fourth set of symbols that encodes a checksum for the y coordinate.

[0090] One alternative is shown in Figure 4C, which is a schematic diagram showing how a fiducial marker (left) may encode polar coordinate (R and θ) information. The marker grid (right) shows the layout of markers (represented as stars) on a substrate. In such cases, the at least one set of symbols may comprise a first set of symbols that encode a distance from the centroid of the spine and a second set of symbols that encode an angle from a reference direction defined by the spine. The reference direction may be defined by the long axis of the spine. In these cases, the at least one set of symbols may comprise a third set of symbols that encode a checksum on the distance and a fourth set of symbols that encode a checksum on the angle.

[0091] These fiducial markers may encode simple instructions for machines in a fully or partially automated manufacturing process. Two examples are described below.

[0092] 4D is a schematic diagram showing how a reference marker may encode coordinate information regarding a position offset from a marker. In this example, each marker may encode coordinate information of an offset relative to the center of each marker (e.g., a marker may encode "Δx, Δy" and the marker has its center on the wafer at position "x0, y0". Thus, the reference position of the marker is (x0+Δx), (y0+Δy)). A machine vision system may scan the substrate for all the markers, which may be used to instruct a machine where to go to each marker, for example for a calibration measurement or for placement of a component in a manufacturing process.

[0093] In these cases, the at least one set of symbols may comprise a first set of symbols encoding a first coordinate of an offset location relative to the centroid of the spine and a second set of symbols encoding a second coordinate of the offset location relative to the centroid of the spine. In such cases, the first and second coordinates may be Cartesian or polar coordinates. The at least one set of symbols may comprise a third set of symbols encoding a checksum for the first coordinate and a fourth set of symbols encoding a checksum for the second coordinate.

[0094] 4E is a schematic diagram showing how fiducial markers may encode instructions for a machine. In this example, each marker may encode instructions as to which component or layout should be placed at a particular location relative to each marker (e.g., a marker may encode "i1, i2..." to instruct the machine to place component i at an offset Δy relative to that marker). Thus, the first and second coordinates may form instructions for a machine in a calibration or manufacturing process.

[0095] Figure 5 shows an example LithoTag design (left) and what it would look like (right) when formed using nanofabrication techniques. LithoTags as small as 850 nm were fabricated as evaporated titanium gold on a silicon substrate. The LithoTag can be seen in the image on the right to show no signs of shape change during fabrication, demonstrating that LithoTags can be detected even at the nanoscale. There is some bridging between some of the features in the LithoTag shown in Figure 5, but it is very slight and does not reduce detection accuracy.

[0096] In general, LithoTags can be fabricated using any suitable technique. For example, LithoTags can be fabricated using deposition or sputtering of metal or dielectric materials onto a substrate (such as a silicon wafer). Thus, LithoTags can be fabricated using additive processing techniques such as lift-off. In another example, LithoTags can be fabricated by etching into a wafer or metal / dielectric layer. Thus, LithoTags can be fabricated using subtractive processing techniques such as etching. In another example, LithoTags can be fabricated by patterning a fluorescent layer. This can be useful in the case of 2D semiconductors, quantum dots, sputtered semiconductors, resulting in semi-transparent markers. In this case, a marker material can be chosen that has a different emission wavelength than the sample to be detected (e.g., quantum dots). In this case, these markers can be distinguished from these samples during fluorescence imaging.

[0097] LithoTag can be used for nanomaterials imaging, materials characterization, device fabrication, semiconductor device or wafer inspection, bio-imaging or fluorescence imaging, image stabilization, precise mapping or stitching, and automatic positioning from any image, etc.

[0098] Figures 6A-6D show various ways in which LithoTags can be made and imaged. In Figure 6A, LithoTags are formed using lift-off onto a substrate (e.g., a silicon wafer) and detected by imaging the substrate. In Figure 6B, LithoTags (opaque) are formed on a transparent substrate and detected by light transmitted through the substrate. In Figure 6C, LithoTags are encapsulated within the sample under investigation. In Figure 6D, LithoTags are encapsulated on a microscope cover slip.

[0099] Accordingly, the present technology further provides an apparatus comprising a substrate and at least one machine-readable reference marker of the type described above in relation to Figures 3A-5, the at least one machine-readable reference marker comprising a spine defining a shape of the marker having a perimeter-to-area ratio of the shape within a predetermined threshold range and a spine defining at least two regions of the marker, and at least one set of symbols provided in one or more of the at least two regions encoding coordinate information of the centroid of the spine on the substrate.

[0100] The at least one machine-readable fiducial marker of the device may be an array of fiducial markers.

[0101] The substrate of the apparatus may be any one of a patternable material, a semiconductor device, a semiconductor wafer, a metal layer, a dielectric layer, a transparent substrate, and a microscope cover slip.

[0102] In some cases, the at least one machine-readable fiducial marker may be formed by patterning the substrate with a contrast material, it will be appreciated that the fiducial markers may be formed using any suitable manufacturing process or technique.

[0103] To read the LithoTags, the present technology provides several detection methods, which are described below with reference to Figures 7-11.

[0104] 7 shows a flow diagram of example steps for identifying machine-readable fiducial markers (i.e., LithoTags) of the type described herein. The method includes acquiring an image of a substrate having machine-readable fiducial markers formed thereon (step S100), acquiring information about properties of the fiducial markers formed on the substrate (step S102), and using the acquired information to identify the fiducial markers in the image (step S104).

[0105] In one example, obtaining information about a characteristic of a fiducial marker formed on the substrate (S102) may include obtaining a perimeter-to-area ratio of a shape defined by a spine of the fiducial marker.

[0106] In this case, the method may identify fiducial markers in the image by using the acquired perimeter-to-area ratio of the shape defined by the spine. FIG. 8 shows example steps for identifying fiducial markers when the perimeter-to-area ratio is known. Thus, if this information is ratio information acquired in step S102, step S104 may include the steps shown in FIG. 8. The steps shown in FIG. 8 are also known as "region feature recognition algorithms." Thus, the method may further include the steps of identifying potential fiducial markers in the image (step S200), determining the centroid of the potential fiducial marker (step S202), using the centroid to determine the major and minor axes of the spine of the potential fiducial marker (step S204), and using the determined major and minor axes to calculate the perimeter-to-area ratio of the potential shape defined by the spine (step S206). The method may include the step of determining whether the calculated perimeter-to-area ratio matches the ratio in the acquired information (step S208). If the calculated perimeter-to-area ratio is determined to match the ratio in the acquired information, the potential fiducial marker is identified as a fiducial marker (step S210).

[0107] The identified fiducial markers can then be used to provide position and / or orientation information of other nearby objects on the substrate. Thus, the method may further include determining the orientation of the fiducial markers using the major and minor axes. Furthermore, the method may include determining fixed positions of at least one set of symbols encoding coordinate information using the acquired information, identifying whether any symbols are present at the fixed positions based on these fixed positions of the symbols, and reading the x and y coordinates of the centroid of the fiducial markers based on the identification of any symbols. In the case where the at least one set of symbols further encodes checksums for the x and y coordinates, the method may include using the symbols to check whether these x and y coordinates have been read correctly.

[0108] If, in step S208, it is determined that the calculated ratio does not match the perimeter to area ratio in the acquired information, then the potential fiducial marker is determined to not be a fiducial marker (step S212). The process for identifying markers may then move on to identifying another potential fiducial marker and repeating the steps shown in Figure 8 with respect to this potential marker. Alternatively, the process may continue with the steps shown in Figure 9.

[0109] Thus, the detection process described with reference to Figure 8 uses the characteristics of the markers to identify the markers on the substrate. Figure 9 illustrates this detection process with reference to potential reference markers on the substrate.

[0110] In Figure 9, image (a) shows an image of potential fiducial markers on a substrate. These fiducial markers were fabricated as described above with reference to Figure 5. This image may be captured using any suitable imaging or microscopy technique, such as, but not limited to, an optical microscope, an electron microscope, or a scanning probe microscope.

[0111] The captured image may be processed before the method for identifying the fiducial markers is performed. This pre-processing may make the identification method easier to perform. Pre-processing may include adjusting the image levels. This level adjustment may correct or adjust the tonal range and color balance of the image, which may facilitate viewing of the fiducial markers. The image may also be converted to an 8-bit grayscale image. Image (b) of FIG. 9 shows the captured image after it has been level adjusted and converted to 8-bit grayscale.

[0112] A Gaussian filter, for example with a standard deviation of 1 pixel, may then be used to remove some or most of the noise in the image. Image (c) of FIG. 9 shows the effect of applying this Gaussian filter. The image is then binarized by selecting a cutoff threshold for pixel values, see image (d) of FIG. 9. Furthermore, this preprocessing may include eroding the edges of the detection regions to remove more noise, for example by selecting a square structuring element (i.e. a matrix defining the neighborhood of each pixel) with a size of 4 pixels in all directions, as shown in image (e). These detection regions are regions with pixel values ​​that lie on the relevant side of the cutoff threshold selected during the binarization step.

[0113] Then, all artifacts in the image that are below a threshold size are removed. This is shown in image (f). This is to ensure that other objects in the image do not interfere with the reading of the fiducial markers. This threshold size may be, for example, 15 pixels. It is not necessary to remove features / artifacts from the image in order for LithoTags to be detected, since the detection is robust to the presence of such features / artifacts if they fall outside a defined circularity limit. A perfect circle has a circularity of 1, whereas for LithoTags the circularity limit is greater than 1.

[0114] The LithoTag features are then expanded, for example by selecting a square structuring element (i.e., a matrix that defines a pixel's neighborhood, e.g., adjacent pixels within a square region of a certain size around the pixel) with a size of 4 pixels in all directions, thereby restoring the original LithoTag as much as possible, see image (g).

[0115] Images (a)-(g) show some pre-processing steps that may be performed before LithoTag detection begins, while images (h) and (i) show how LithoTag detection is performed.

[0116] In image (h), potential fiducial markers are identified in the image, where the box shows the outline of the detected tag and the number shown in the image is the circularity of the detected tag. This may involve identifying regions in the image where the pixel value is not zero and the perimeter to area ratio or circularity is within a certain threshold range. If the tag spines are highly circular, this means that most of the region remaining after the above processing steps is expected to be LithoTags, even in images that contain a high percentage of nanowires or other high aspect ratio features. Removal of any remaining outliers requires ensuring that the ratio of the second moment around the major and minor axes is within an acceptable range of expected values, and that the percentage of detected region is within an acceptable range of expected values.

[0117] Image (i) illustrates a method using characteristics of a potential fiducial marker to determine whether an actual fiducial marker has been identified. The centroid of the potential fiducial marker is determined, which is shown in image (i) as a "plus" sign. The centroid is used to establish a local coordinate frame for the spine of the potential fiducial marker, the local coordinate frame having a major axis corresponding to the y-axis and a minor axis corresponding to the x-axis. The determined major and minor axes are used to calculate a perimeter-to-area ratio of the shape defined by the spine. The method may include determining whether the calculated perimeter-to-area ratio matches a known perimeter-to-area ratio of the LithoTag, and identifying the potential fiducial marker as a fiducial marker if the calculated perimeter-to-area ratio is determined to match the known ratio.

[0118] Once the affine transformation (rotation, translation, and scale) of the LithoTag is known, the fixed location of the binary marker can be easily found and it can be determined whether the binary marker is set or not. If present, the LithoTag identification / reading method may include calculating checksums of the x and y coordinates and checking that they match the checksum values ​​in the LithoTag.

[0119] Returning to FIG. 7, in one example, obtaining information about a characteristic of a fiducial marker formed on the substrate (S102) may include obtaining a template of a spine of the fiducial marker.

[0120] In this case, the method can use the acquired template to identify a fiducial marker in the image, or to identify any potential fiducial markers that could not be properly read or identified using the process shown in FIG. 8. FIG. 10 shows example steps for identifying a fiducial marker when its template is known. Thus, if this information was acquired in step S102 of FIG. 7, step S104 may include the steps shown in FIG. 10. The method may further include identifying a potential fiducial marker in the image (step S300), rotating and scaling the template to match the orientation and size of the potential fiducial marker (step S302), and calculating a correlation value between the identified potential fiducial marker in the image and the template (step S304). The method may include determining whether the calculated correlation value indicates that the potential fiducial marker closely matches the template (step S306). If the calculated correlation value indicates a match, the potential fiducial marker is identified as a fiducial marker (step S308).

[0121] The identified fiducial markers can then be used to provide position and / or orientation information for other nearby objects on the substrate. Thus, the method may further include using the template to identify whether any symbols are present in place, and reading the x and y coordinates of the centroids of the fiducial markers based on the identification of any symbols. If at least one set of symbols further encodes a checksum for the x and y coordinates, the method may include using the symbols to check whether the x and y coordinates have been correctly read.

[0122] If the calculated correlation value is determined not to indicate a match in step S306, then the potential reference marker is determined not to be a reference marker (step S310). This process for identifying markers may then move on to identifying another potential reference marker and repeating the steps shown in FIG. 10 with respect to this potential marker.

[0123] Thus, the detection process described with reference to Figure 10 uses templates to identify markers on the substrate. This detection process may allow for processing of noisier images with small features at the expense of some of the speed and flexibility of the other detection processes (Figure 8). Figure 11 illustrates this detection process with reference to potential fiducial markers on the substrate.

[0124] FIG. 11 shows the detection process (also called the "correlation algorithm") that uses an input template of LithoTag spines to detect LithoTags on a substrate. The template of the fiducial marker spines can be obtained in two ways. For example, the template (which may include scale and orientation information) may be known and thus input into the process. Alternatively, the template may not be known and needs to be known by fitting potential templates of different rotations and sizes to several images to find the optimum value of the template. This template is then input to detect other LithoTags.

[0125] The correlation algorithm involves identifying potential fiducial markers in an image. See image (a) of FIG. 11. A template of LithoTag spines (image (b)) is then rotated and scaled to approximately the same orientation and scale as the LithoTags in the image of the potential fiducial markers. That is, the algorithm involves rotating and scaling the template to match the orientation and size of the potential fiducial markers. The algorithm may then include calculating a correlation value (or cross-correlation value) between the identified potential fiducial markers in the image (i.e., image (a)) and the template (image (b)). This cross-correlation value is an indication of the similarity between the potential fiducial markers and the template. Any suitable correlation / cross-correlation analysis or function may be used. This cross-correlation process may result in a value indicative of the similarity of the potential fiducial markers and the template. Alternatively, the cross-correlation process may result in a numerical value per pixel. In this case, pixels in the image (image (a)) that correspond to the template will result in a higher numerical value than pixels in the image that do not correspond to the template. Thus, a high correlation value or peak may correspond to a LithoTag. With an appropriate transformation, it may be possible to determine the exact coordinates of the LithoTag in image (a), see image (c). (For example, a geometric transformation using two sets of coordinates may be generated and used to transform any selected coordinates). Once the affine transformation (rotation, translation, scale) of the LithoTag is known, the fixed position of the binary marker can be easily found and it is possible to determine whether the binary marker is set or not. If present, the LithoTag identification / reading method may include calculating checksums of the x and y coordinates and checking that they match the checksum values ​​in the LithoTag.

[0126] As previously mentioned, if the markers are identified using the process shown in Figure 8 or Figure 10, the method may further include using the acquired information to identify any symbols that encode a checksum for an x ​​coordinate and a checksum for a y coordinate, reading these checksums, calculating checksums for the x and y coordinates, and checking whether the calculated checksum matches the read checksum, where if the calculated checksum matches the read checksum then the potential fiducial marker has been correctly identified and read.

[0127] In both methods described above with reference to Figures 8 and 10, an image of the substrate or sample is acquired and then potential fiducial markers are identified in the image for further analysis. To do this, it is necessary to identify areas of interest in the image of the substrate or sample. That is, if the substrate contains nanowires, it may be necessary to identify the location of these nanowires before identifying potential fiducial markers located in the vicinity of the nanowires. Thus, the methods described above may include a further process for identifying areas of the image of interest.

[0128] The process of identifying the region of interest may include maximizing the contrast between the nanofeatures of interest and the substrate using image processing. This process may include searching for objects and filtering based on object properties. This image processing may include converting the image to grayscale, adjusting the intensity values ​​in the grayscale image to maximize contrast, removing small regions below a threshold size from the image, and morphologically closing the image by expanding the features using a square structuring element. After sufficient contrast is achieved, the region of interest is located, such as by using the "regionprops" command in Matlab.

[0129] Regions of interest can be located, for example, using parameter filtering. Parameters considered can include, but are not limited to, region area, length, width, orientation, circularity, solidity, maximum intensity, and minimum intensity. The particular parameters used and their values ​​can be tailored to the particular application and features of interest to be identified. For example, nanowire identification can use aspect ratio based filtering. Parameter filtering in this case can be tailored to recognize nanowires that are non-overlapping and have particular dimensions. Similarly, triangular transition metal dichalcogenide (TMD) flakes can be detected based on circularity and length. Furthermore, intensity based filtering can distinguish between different layers of TMDs, with single layer flakes being brighter than multilayer flakes.

[0130] As described above, the present technology may provide an apparatus that includes a substrate and at least one machine-readable fiducial marker. In some cases, the substrate may include an array or grid of fiducial markers. A grid of any size may be formed on the substrate. For example, a 16x16 (for 4-bit LithoTags) or 512x512 (for 9-bit LithoTags) grid may be formed on a silicon substrate to allow for high-precision positioning. Creating a LithoTag grid may require specifying the number of LithoTags in the x and y directions, the size of each LithoTag, and the spacing between adjacent LithoTags. For example, the spacing between adjacent LithoTags may be expressed as the distance between the centroids of adjacent LithoTags, or the distance between any two other specific points of the two LithoTags, where both of these two points are the same distance away from the centroids of the LithoTags.

[0131] Additionally, the present technology provides a method for using an array of machine-readable fiducial markers of the type provided herein on a substrate to control a machine during a machine calibration or manufacturing process. The method includes capturing at least one image of the array of machine-readable fiducial markers, identifying at least one machine-readable fiducial marker in the array using a method provided herein, and using the identified at least one fiducial marker to control the machine. Some examples are now described.

[0132] Estimating translation stage position. Many translation stages are not well calibrated and the actual position may differ from the indicated position. This is especially true if the stage moves over a large distance. One solution is to track the stage position, for example by incorporating a laser interferometer into the translation stage. This solution can be expensive, but most of the more affordable translation stages suffer from some form of drift.

[0133] The spacing of the fiducial markers is known to be within manufacturing tolerances (a few nm) and the positions of the fiducial markers are encoded in the fiducial markers themselves. This can be used to infer the actual position of the translation stage, thereby correcting for any drift even when using an uncalibrated stage. For example, the translation stage is moved to an initial position and an image is captured. The fiducial markers in the image are used to determine that the center of this image is located at position (5,5) in the marker coordinate system. The translation stage is then moved along the x-axis to the indicated position of 500 μm, and then the center of the image is determined to be located at position (10.1,5) in the marker coordinate system. As the markers are known to be spaced 100 μm apart, the actual distance traveled by the translation stage is calculated as 510 μm, thus allowing a calibration factor of 500 / 510 to be set for the x-axis movement.

[0134] The motion of the translation stage can be non-linear, especially over larger distances. The fiducial markers can be used for feedback to correct for this non-linearity by using an iterative loop. For example, an image is captured and the center of this image is determined to correspond to position (5, 5) in the marker coordinate system. It is desired to move 1000 μm in the positive x direction. The calibration coefficients have been determined to be 500 / 510 for this translation stage, so the translation stage is commanded to move 1000*500 / 510 μm=980 μm. The translation stage has a non-linear response and the previously determined calibration coefficients are not valid over this range. When a new image is captured, it can be seen that the center of the image is located at the current position (14.9, 5). A further small move of 10*500 / 510 μm=9.8 μm can be made to get closer to the target position. If necessary, multiple small moves can be made to get asymptotically closer to the target position.

[0135] The control method may thus be for calibrating a translation stage of an imaging machine, where the method may include reading an identified fiducial marker to determine first and second coordinates of a position relative to a centroid of the fiducial marker, moving the translation stage a predetermined distance, determining a distance traveled by the translation stage relative to the identified fiducial marker using an array of machine readable fiducial markers, calculating an actual distance traveled by the translation stage, and calibrating the translation stage using the actual distance.

[0136] Estimating Feature Size. When an image is captured with a microscope, the scale of the features is determined by several factors, such as the choice of lens. While it is possible to calculate the magnification of the microscope, it is often more reliable to image a calibration target made up of features with known sizes. These fiducial markers can themselves serve as calibration targets, as they are known to be sized and spaced within manufacturing tolerances.

[0137] For example, a substrate is used that is patterned with fiducial markers spaced 100 μm apart from each other. An image is taken capturing the fiducial markers. In this image, the fiducial markers are spaced 1000 pixels apart. As a result, it is determined that one pixel is approximately 0.1 μm. This allows a scale bar to be placed in the image with a width of 500 pixels, which represents a length of 50 μm. Additionally, a nanowire is present in the image with a width of 205 pixels. In this case, it is possible to calculate that the length of the nanowire is 20.5 μm.

[0138] The method may thus be a method for calibrating the magnification of an imaging machine, comprising using an array of machine-readable fiducial markers to determine the spacing between the fiducial markers in pixels, and calculating the magnification of the imaging machine using the known distance between the fiducial markers in the array.

[0139] Focus Feedback: It can sometimes be difficult to determine if an image is perfectly in focus. Furthermore, it can be desirable to set the focus automatically, for example when taking multiple images of a large substrate, and when the focus position can change with xy position.

[0140] These fiducial markers can be used to provide feedback on the degree of focus of the image, given the above template / convolution approach to detecting fiducial markers. In this example, the microscope image is convolved with the fiducial marker template image, and any peaks in the convolution indicate the presence of a fiducial marker. The magnitude of these peaks will be higher if the microscope image is more in focus, providing a useful feedback mechanism. This convolution approach takes advantage of the unique shapes of fiducial markers that are unlikely to exist in nature. Additionally, this approach takes advantage of the fact that when imaging a fiducial marker substrate, there will always be a fiducial marker in the field of view.

[0141] For example, a large 2" x 2" wafer is imaged using raster scanning. The wafer is focused at the lower left corner before the start of the scan. As the scan progresses, it is found that the wafer is out of focus, so it would be useful to automatically adjust the focus. The scan is temporarily stopped. For a starting height of z=0 μm, images are acquired at 11 different heights, from z=-5 μm to z=+5 μm. Each of these images is convolved with the fiducial marker template. The magnitude of the maximum peak in each convolved image is measured and it is found that the image at z=+3 μm has the largest peak. This indicates that the image at z=+3 μm is most likely in focus. The translation stage is set to a height of z=+3 μm and the raster scan continues.

[0142] The method may thus be for calibrating the focus of an imaging machine, in which case the method may include convolving the captured image with a template of a spine of a fiducial marker, identifying a peak in the convolution indicative of the presence of the fiducial marker, and controlling the imaging machine to adjust the focus to raise the peak in the convolution.

[0143] Image Stitching: A common metrology is imaging a large area, e.g. an entire wafer. Given the limited field of view of the microscope, this is done by raster scanning, i.e. taking an image, moving the translation stage, taking another image, etc.

[0144] After performing the raster scan, it is often desirable to stitch the acquired images together. This stitching can be difficult to do correctly because most translation stages do not have ideal repeatability and are not perfectly calibrated. One solution is to use a stage with an interferometer. Another, more affordable solution is to use feedback from the images. Typically, the acquired images are considered to overlap to some extent, and then the edges of the images are compared and aligned based on features visible in the overlap. The comparison and alignment relies on identifiable and recognizable features located in the edge regions, and therefore may not always work. When using substrates patterned with fiducial markers, each image needs to capture at least one fiducial marker. If each fiducial marker is encoded with position information, it is possible to determine the exact position of each image relative to any other image, which can then be used to stitch the images together. Importantly, this stitching does not rely on the uniqueness of other features (which the fiducial markers are already unique), nor does it rely on overlap between the images.

[0145] Multi-scale imaging. Figure 12 is an image of an array with fiducial markers of two different sizes. A substrate is patterned with a grid of fiducial markers 200, 202. Every fifth fiducial marker 202 in the grid is larger (e.g., 10 μm total width) and therefore detectable with an optical microscope. All other fiducial markers 200 are small (e.g., 2 μm total width) and therefore detectable with a scanning electron microscope. As a result, it is still possible to image the substrate and detect the fiducial markers in both SEM and optical microscopes. That is, an area of ​​interest seen in one microscope can be easily relocated in the other microscope.

[0146] Multimodal imaging: These fiducial markers can be detected optically, but also with other instruments, such as AFM or imaging ellipsometers. As a result, areas of interest found in one instrument can be easily relocated in another instrument.

[0147] For example, a nanowire of interest is found in an SEM. A nearby fiducial marker is then detected, allowing a determination to be made that the nanowire is located at (5.5, 5.5) in the marker coordinates. The substrate is then inserted into a standard optical microscope, where the fiducial marker grid allows easy relocation of the (5.5, 5.5) position, allowing the nanowire of interest to be easily found in a position that allows both dark field and bright field imaging. The substrate is then inserted into a Raman microscope, where the fiducial marker again allows relocation of the nanowire at (5.5, 5.5). Finally, the substrate is inserted into an AFM, where the fiducial marker grid again allows easy relocation of the nanowire at (5.5, 5.5).

[0148] Although what is believed to be the best mode of carrying out the present technology and other modes as appropriate have been described above, those skilled in the art will readily appreciate that the present technology should not be limited to the specific configurations and methods disclosed in the description of the preferred embodiments. Those skilled in the art will recognize that the present technology has broad applications and that these embodiments may undergo wide modifications without departing from any inventive concept defined in the appended claims. [Explanation of symbols]

[0149] 100 Fiducial Markers 102 areas 104 areas 106 areas 108 areas 110 Spine 200 fiducial markers 202 fiducial marker

Claims

1. A machine-readable fiducial marker, comprising: a spine having a minor axis and a major axis, the orientation of the major axis and the minor axis defining an orientation of the marker on a substrate; a shape of the marker having a perimeter-to-area ratio within a predetermined threshold range; and At least two regions of the marker a spine defining the at least one set of symbols disposed in one or more of the at least two regions, the at least one set of symbols encoding coordinate information of a location relative to a centroid of the spine on a substrate; A fiducial marker comprising:

2. The fiducial marker of claim 1 , wherein the perimeter-to-area ratio defined by the spine indicates the circularity of the shape.

3. 3. The fiducial marker of claim 1 or 2, wherein the at least one set of symbols comprises a first set of symbols encoding an x-coordinate of the centroid of the spine and a second set of symbols encoding a y-coordinate of the centroid of the spine.

4. 4. The fiducial marker of claim 3, wherein the at least one set of symbols comprises a third set of symbols that encodes a checksum for the x coordinate and a fourth set of symbols that encodes a checksum for the y coordinate.

5. 3. The fiducial marker of claim 1 or 2, wherein the at least one set of symbols comprises a first set of symbols that encode a distance from the centroid of the spine and a second set of symbols that encode an angle from a reference orientation defined by the spine.

6. 4. The fiducial marker of claim 3, wherein the at least one set of symbols comprises a third set of symbols encoding a checksum on the distance and a fourth set of symbols encoding a checksum on the angle.

7. 3. The fiducial marker of claim 1 or 2, wherein the at least one set of symbols comprises a first set of symbols encoding a first coordinate of an offset position of the spine relative to the centroid and a second set of symbols encoding a second coordinate of an offset position of the spine relative to the centroid.

8. The fiducial marker of claim 7 , wherein the first coordinate and the second coordinate are Cartesian coordinates.

9. The fiducial marker of claim 7 , wherein the first coordinate and the second coordinate are polar coordinates.

10. 10. The fiducial marker of claim 7, 8 or 9, wherein the at least one set of symbols comprises a third set of symbols encoding a checksum for the first coordinate and a fourth set of symbols encoding a checksum for the second coordinate.

11. 11. A reference marker according to any one of claims 7 to 10, wherein the first and second coordinates form instructions for a machine in a calibration or manufacturing process.

12. 12. The fiducial marker of claim 1, wherein the spine comprises at least two arms.

13. 13. The fiducial marker of claim 12, wherein the spacing between each symbol in at least one set of symbols is the same as the width of each arm of the spine.

14. 14. The fiducial marker of claim 1, wherein the spine defines a hexagonal marker, a square marker, or a rectangular marker.

15. 15. A fiducial marker according to any preceding claim, wherein at least one set of symbols encoding coordinate information are arranged in a predetermined pattern indicating a fixed position of each symbol.

16. 16. The fiducial marker of claim 15, wherein the predetermined pattern indicating the fixed position of each symbol consists of 4, 9, or 16 positions.

17. the spine is a cruciform spine having four arms and defining four regions of the marker; At least one set of symbols comprises: a first set of symbols disposed within a first region of the four regions, the first set of symbols being arranged in a predetermined pattern indicating a fixed position for each symbol and forming a binary representation of first coordinates of the position relative to the centroid of the spine on the substrate; a second set of symbols disposed within a second one of the four regions, the second set of symbols being arranged in a predetermined pattern indicating a home position of each symbol and forming a binary representation of second coordinates of the location relative to the centroid of the fiducial marker on the substrate; a third set of symbols in a third region of the four regions, the third set of symbols defining a checksum for the first coordinate; a fourth set of symbols in a fourth region of the four regions, the fourth set of symbols defining a checksum for the second coordinate; and 17. The fiducial marker of any one of claims 1 to 16, comprising:

18. An apparatus comprising: A substrate; at least one machine-readable fiducial marker, a spine having a minor axis and a major axis, the orientation of the major axis and the minor axis defining an orientation of the marker on a substrate; a shape of the marker having a perimeter-to-area ratio within a predetermined threshold range; and At least two regions of the marker A spine defining the At least one set of symbols provided in one or more of the at least two regions, the at least one set of symbols encoding coordinate information of a location on a substrate relative to a centroid of the spine. at least one machine-readable fiducial marker comprising: An apparatus comprising:

19. 20. The apparatus of claim 18, wherein the at least one machine-readable fiducial marker is an array of fiducial markers.

20. 20. The apparatus of claim 19, wherein the array of fiducial markers comprises fiducial markers of at least two different sizes.

21. 21. The apparatus of any one of claims 18 to 20, wherein the substrate is any one of a patternable material, a semiconductor device, a semiconductor wafer, a metal layer, a dielectric layer, a transparent substrate, and a microscope cover slip.

22. 22. The apparatus of claim 21, wherein the fiducial markers are formed by patterning the substrate with a contrast material.

23. 18. A method for identifying machine-readable fiducial markers of the type according to any one of claims 1 to 17 on a substrate, comprising the steps of: acquiring an image of a substrate on which the machine-readable fiducial markers are formed; obtaining information regarding a characteristic of the fiducial marker formed on the substrate; using the acquired information to identify fiducial markers in the image; A method comprising:

24. 24. The method of claim 23, wherein obtaining information regarding a characteristic of the fiducial marker comprises a perimeter-to-area ratio of a shape defined by the spines of the fiducial marker.

25. identifying potential fiducial markers in the image; determining the centroid of the potential fiducial marker; determining major and minor axes of a spine of the potential fiducial marker using the centroid; calculating a perimeter to area ratio of a shape defined by the spine using the determined major and minor axes; determining whether the calculated perimeter to area ratio matches the ratio in the acquired information; identifying the potential fiducial marker as a fiducial marker if the calculated perimeter-to-area ratio is determined to match the ratio in the acquired information.

25. The method of claim 24, further comprising:

26. Determining an orientation of the fiducial marker using the major and minor axes.

26. The method of claim 25, further comprising:

27. using the acquired information to determine fixed locations of at least one set of the symbols encoding coordinate information; identifying whether any symbol is present at the fixed position based on the fixed positions of the symbols; reading first and second coordinates of a position of the fiducial marker relative to the centroid based on an identification of an arbitrary symbol; 27. The method of claim 25 or 26, further comprising:

28. 28. The method of claim 23, wherein obtaining information about a characteristic of the fiducial marker comprises a template of the spine of the fiducial marker.

29. identifying potential fiducial markers in the image; rotating and scaling the template to match the orientation and size of the potential fiducial markers; calculating a correlation value between the identified potential fiducial markers in the image and the template; identifying the potential reference marker as a reference marker if the correlation value indicates that the potential reference marker closely matches the template; 30. The method of claim 28, further comprising:

30. using said template to identify whether any symbols are present at said fixed positions; reading first and second coordinates of a location of the fiducial marker relative to the centroid based on the identification of any symbol; 30. The method of claim 29, further comprising:

31. using said acquired information to identify any symbols encoding a checksum for said first coordinate and a checksum for said second coordinate and reading both said checksums; calculating a checksum for the first coordinate and the second coordinate; checking whether the calculated checksum matches the read checksum; Further comprising:

31. The method of claim 27 or 30, wherein if the calculated checksum matches the read checksum, then the potential reference marker has been correctly identified and read.

32. 18. A method for controlling a machine during a machine calibration or manufacturing process using an array of machine-readable reference markers of the type claimed in any one of claims 1 to 17 on a substrate, comprising the steps of: capturing at least one image of said array of machine-readable fiducial markers; Identifying at least one machine-readable fiducial marker in the array using the method of any one of claims 23 to 31; controlling a machine using said identified at least one fiducial marker; A method comprising:

33. The method is for calibrating a translation stage of an imaging machine, reading the identified fiducial marker to determine first and second coordinates of a location of the fiducial marker relative to the centroid; moving the translation stage a predetermined distance; using the array of machine-readable fiducial markers to determine a distance traveled by the translation stage relative to the identified fiducial marker; calculating an actual distance traveled by the translation stage; calibrating the translation stage using the actual distance; 33. The method of claim 32, comprising:

34. The method is for calibrating a magnification of an imaging machine, using the array of machine-readable fiducial markers to determine spacing between the fiducial markers in pixels; calculating the magnification of the imaging machine using the known distances between the fiducial markers in the array; 33. The method of claim 32, comprising:

35. The method is for calibrating a focus of an imaging machine, convolving the captured image with a template of the spine of the fiducial marker; identifying peaks in said convolution that are indicative of the presence of fiducial markers; controlling the imaging machine to adjust the focus to increase the peak in the convolution; 33. The method of claim 32, comprising:

36. A non-transitory data carrier carrying code which, when implemented in a processor, causes the processor to carry out the method of any one of claims 23 to 31 and any one of claims 32 to 35.

Images