Data processing device, data processing system, data processing method, and program

Abstract

The invention relates to a data processing apparatus, a data processing system, a data processing method, and a program. A data processing device (3) is provided with: a receiving means (35) for receiving measurement data from a hyperspectral device (2) capable of measuring spectral data for each pixel of an image of a measurement object (4); a detection unit (36) that detects the occurrence of defects in the received measurement data; and a file generation means (37) that, when the occurrence of a defect in the measurement data is detected, generates a measurement data file in which a specifying value capable of specifying the defect data is embedded in a site where the defect occurs.

Description

Technical Field

[0001] The present invention relates to a data processing device, a data processing system, a data processing method and a program for generating a measurement data file based on measurement data of a hyperspectral device capable of measuring two-dimensional spectroscopic data of a measurement object. Background Art

[0002] In the data processing system using a hyperspectral device such as a hyperspectral camera capable of measuring two-dimensional spectral data as described above, a measurement data file summarized as a certain unit is generated based on the measurement data of the HSI device and stored in a storage unit, and data is read from the stored file to perform image generation, spectral data analysis, etc. In the following description, a "hyperspectral camera" is sometimes described as an "HSI camera", and a "hyperspectral device" is sometimes described as an "HSI device".

[0003] By HSI equipment Figure 8 The measurement data when measuring the measurement object 100 shown in (A) is a spectral data 102 corresponding to the number of wavelengths based on the measurable wavelength range and the resolution. Figure 8 (C)), thus, it is different from the data 101 obtained by a conventional RGB 3-channel camera ( Figure 8 Compared with (B), the amount of data is quite large.

[0004] In addition, for example, in a measurement object such as a micro LED wafer in which a large number of LED elements with a size of tens of μm are produced on a substrate with a diameter of about 100 mm, it is sometimes required to analyze and evaluate the spectral characteristics of each of these elements during inspection.

[0005] In order to achieve such a requirement, it is necessary to perform measurements at a high resolution at high speed. However, depending on the hardware specifications of the data processing device side such as a personal computer (hereinafter also recorded as PC) that receives the measurement data, specifically the performance of the data receiving unit, storage unit (SSD, HDD, etc.), etc., it is considered that the storage of the measurement data may not keep up with the measurement.

[0006] In this case, data loss occurs in which measurement data is omitted. Conventionally, the missing data is interpolated using data around the missing portion, and a measurement data file is generated as if no data loss occurs.

[0007] However, when generating a measurement data file by interpolating missing data, there are the following problems.

[0008] That is, after the measurement data file is generated, it is sometimes necessary to perform post-processing using the generated measurement data file. In this case, depending on the processing content, it is sometimes necessary to explicitly indicate the location to be interpolated (the location where data loss occurs) or to ignore the data. However, if the missing data has already been interpolated, there is a problem that these post-processing will be hindered.

[0009] In particular, the measurement data of a hyperspectral device is two-dimensional spectroscopic data, so a wide range of analytical processing is performed using the measurement data file. If interpolation has already been completed, these analytical processing is restricted, which prevents the execution of appropriate processing.

[0010] In addition, Patent Document 1 proposes an image processing device that can repair defective areas at high speed and high quality for an input image including defective areas such as punching holes and marks of a binding machine. The image processing device includes: a texturing unit that texturizes pixel values ​​of pixels in an area that does not include the defective area; and a repairing unit that performs image processing to infer pixel values ​​of pixels included in the defective area based on the texturized pixel values ​​and repair the defective area.

[0011] Patent Document 1: Japanese Patent Application Publication No. 2011-35567

[0012] However, the image processing device described in Patent Document 1 is not a technology related to data loss of measurement data from an HSI device having a large amount of measurement data, and therefore it is difficult to apply to measurement data from an HSI device. Therefore, according to Patent Document 1, the above-mentioned problem that the post-processing using the measurement data file cannot be properly performed when the measurement data file is generated in a state where the defective data is interpolated cannot be solved. Summary of the invention

[0013] The present invention has been made in view of such technical problems, and its object is to provide a data processing device, a data processing system, a data processing method and a program that can appropriately perform post-processing using a generated measurement data file when measurement data of an HSI device is defective.

[0014] The above purpose is achieved through the following technical solutions.

[0015] (1) A data processing device comprising:

[0016] A receiving unit receives measurement data from a hyperspectral device capable of measuring two-dimensional spectroscopic data of a measurement object;

[0017] a detection unit configured to detect generation of defects in the measurement data received by the receiving unit; and

[0018] The file generating unit generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a defect-occurring portion when the occurrence of defect in the measurement data is detected by the detecting unit.

[0019] (2) A data processing device according to item 1 above, wherein:

[0020] An analyzing unit is provided, wherein the analyzing unit analyzes the measurement data file generated by the file generating unit,

[0021] The analysis unit switches the operation of the defect occurrence location of the measurement data file according to the purpose of analysis.

[0022] (3) A data processing device according to item 1 or 2 above, wherein:

[0023] The above-mentioned determined value is 0 or the maximum value or the minimum value that can be held with respect to the numerical value of the measurement data.

[0024] (4) A data processing device according to item 2 above, wherein:

[0025] The analysis unit interpolates the measurement data at the defect generation location by using at least any one of linear interpolation, spline interpolation and Lagrange interpolation.

[0026] (5) A data processing device according to item 2 above, wherein:

[0027] The detection unit counts the occurrence of defects and determines whether or not a predetermined number or more of defects have continuously occurred.

[0028] (6) A data processing device according to item 5 above, wherein:

[0029] The object to be measured is a micro LED wafer having a plurality of LED elements, and the predetermined value is the number of times defects occur when an unresolvable area generated due to continuous occurrence of defects exceeds the size of the LED elements.

[0030] (7) A data processing device according to item 5 above, wherein:

[0031] A control unit is provided. When the detection unit determines that defects exceeding a preset specified value are continuously generated, the control unit controls the hyperspectral device to interrupt the measurement performed by the hyperspectral device or re-measure the measurement section including the area where the defects are continuous.

[0032] (8) A data processing device according to item 2 above, wherein:

[0033] The object to be measured is a micro LED wafer having a plurality of LED elements.

[0034] The analysis unit generates an image of the micro LED wafer based on the measured spectral data, and determines that data is missing when the determined value is detected during image generation, and directly uses the determined value or replaces it with another value or uses the previous and next data for interpolation.

[0035] (9) A data processing device according to item 2 above, wherein:

[0036] The object to be measured is a micro LED wafer having a plurality of LED elements.

[0037] The analysis unit performs analysis in units of LED elements on the micro LED wafer, and when the specific value is detected, excludes pixels having the specific value from analysis targets.

[0038] (10) A data processing device according to item 9 above, wherein:

[0039] The analysis unit adds warning information to the analysis result of the LED element when the number of pixels not excluded from the analysis target is less than a predetermined value.

[0040] (11) A data processing system, wherein:

[0041] A hyperspectral device capable of measuring two-dimensional spectral data of a measurement object and a data processing device according to the above item 1 or 2 are provided.

[0042] (12) A data processing method, comprising:

[0043] A measuring step, measuring two-dimensional spectroscopic data of a measuring object by a hyperspectral device;

[0044] a detection step of detecting generation of defects in the measurement data measured by the measurement step; and

[0045] The file generating step generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a location where the defect has occurred, when the occurrence of defect in the measurement data is detected in the detecting step.

[0046] (13) A program for causing a computer to execute:

[0047] A receiving step of receiving measurement data from a hyperspectral device capable of measuring two-dimensional spectroscopic data of a measurement object;

[0048] a detecting step of detecting the occurrence of a defect in the measurement data received in the receiving step; and

[0049] The file generating step generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a location where the defect has occurred, when the occurrence of defect in the measurement data is detected in the detecting step.

[0050] (14) The procedure according to item 13 above, wherein:

[0051] further causing the computer to execute an analysis step, in which the measurement data file generated by the file generation step is analyzed,

[0052] In the analysis step, the computer is caused to execute a process of switching the location where the defect of the measurement data file occurs according to the purpose of analysis.

[0053] (15) A procedure according to paragraph 13 or 14 above, wherein:

[0054] The above-mentioned determined value is 0 or the maximum value or the minimum value that can be held with respect to the numerical value of the measurement data.

[0055] (16) A procedure according to paragraph 13 or 14 above, wherein:

[0056] In the analysis step, the computer is caused to execute the following processing: at the defect occurrence site, the measurement data is interpolated by at least one interpolation method among linear interpolation, spline interpolation and Lagrange interpolation.

[0057] (17) The procedure according to item 14 above, wherein:

[0058] In the detection step, the computer is caused to execute the following processing: the occurrence of defects is counted, and it is determined whether a number of defects exceeding a predetermined value set in advance has been continuously generated.

[0059] (18) The procedure according to item 17 above, wherein:

[0060] The object to be measured is a micro LED wafer having a plurality of LED elements, and the predetermined value is the number of times defects occur when an unresolvable area generated due to continuous occurrence of defects exceeds the size of the LED elements.

[0061] (19) The procedure according to item 17 above, wherein:

[0062] The computer is caused to execute a control step. In the control step, when it is determined by the detection step that defects exceeding a preset specified value are continuously generated, the hyperspectral device is controlled to interrupt the measurement performed by the hyperspectral device or to re-measure the measurement portion of the area including the continuous defects.

[0063] (20) The procedure according to item 14 above, wherein:

[0064] The object to be measured is a micro LED wafer having a plurality of LED elements.

[0065] In the above analysis step, the above computer performs the following processing: generating image data of the above micro LED wafer based on the measured spectral data, and judging that the data is missing when the above determined value is detected during image generation, and directly using the determined value or replacing it with other values ​​or using the previous and next data for interpolation.

[0066] (21) The procedure according to item 14 above, wherein:

[0067] The object to be measured is a micro LED wafer having a plurality of LED elements.

[0068] In the analysis step, the computer is caused to execute a process of performing analysis on each LED element on the micro LED wafer and, when the specific value is detected, excluding pixels having the specific value from the analysis target.

[0069] (22) The procedure according to item 21 above, wherein:

[0070] In the analysis step, the computer is caused to execute a process of adding warning information to the analysis result of the LED element when the number of pixels not excluded from the analysis target is smaller than a predetermined value.

[0071] According to the data processing device, data processing system and data processing method of the present invention, the occurrence of a defect in two-dimensional spectroscopic data of a measurement object measured by an HSI device is detected. When the occurrence of a defect in the measurement data is detected, a measurement data file is generated in which a specific value capable of identifying the defective data is embedded in the location where the defect occurs.

[0072] In this way, a specific value capable of identifying the missing data is embedded in the data missing portion of the generated measurement data file, so when the measurement data file is used for subsequent processing such as analysis processing, the data processing device can recognize that the measurement data is missing. Therefore, the interpolation processing in the subsequent processing can of course perform various operations such as clearly indicating the location where the data is missing or ignoring the data, so that appropriate processing corresponding to the content of the subsequent processing can be performed.

[0073] According to the program of the present invention, a computer can be caused to execute: a receiving step of receiving measurement data from an HSI device capable of measuring two-dimensional spectroscopic data of a measurement object; a detecting step of detecting the occurrence of defects in the measurement data received by the receiving step; and a file generating step of generating a measurement data file having a determination value capable of determining the defective data embedded in the defect occurrence location when the occurrence of the defect in the measurement data is detected by the detecting step. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 This is a block diagram showing the structure of a data processing system according to one embodiment of the present invention.

[0075] Figure 2 This is a diagram schematically showing a part of an image of a measurement object using pixels.

[0076] Figure 3 (A) is a diagram showing a portion of a micro LED wafer as a measurement object. Figure 3 (B) is a diagram schematically showing an image corresponding to a LED element and its surroundings using a plurality of pixels. Figure 3 (C) is a diagram for explaining a state where data loss occurs. Figure 3 (D) is a diagram showing a state where a data defect occurrence portion is filled with a determined value.

[0077] Figure 4 (A) is a diagram schematically showing an image corresponding to an LED element and its surroundings using a plurality of pixels, and is a diagram showing a portion corresponding to an LED element using a grayscale image. Figure 4 (B) is for Figure 4 The grayscale image (A) shows the location where data loss occurs with the maximum value (white). Figure 4 (C) is an explanatory diagram of the case where the generated defective data is interpolated. Figure 4 (D) is a diagram showing the state after interpolation.

[0078] Figure 5 (A) is an explanatory diagram of a case where, when data loss of one line occurs, analysis processing is performed using other pixels where data loss does not occur. Figure 5 (B) is an explanatory diagram of a state where data loss occurs continuously.

[0079] Figure 6 This is a flowchart showing a process of detecting data loss and generating a measurement data file executed by a data processing device.

[0080] Figure 7 This is a flowchart showing the analysis process executed by the data processing device.

[0081] Figure 8 (A) is a diagram showing a measurement object, Figure 8 (B) is a diagram showing data obtained by a RGB3-channel camera. Figure 8 (C) is a graph showing data obtained by a hyperspectral device. DETAILED DESCRIPTION

[0082] Hereinafter, embodiments of the present invention will be described based on the drawings.

[0083] Figure 1 1 is a block diagram showing the structure of a data processing system 1 according to an embodiment of the present invention. The data processing system 1 includes an HSI camera 2 as a hyperspectral device (HSI device) and a data processing device 3 constituted by a personal computer (PC).

[0084] The HSI camera 2 is a well-known structure that measures two-dimensional spectroscopic data of the object to be measured, in other words, data corresponding to the wavelength range and the number of wavelengths based on the resolution that can be measured at each position of the object to be measured, by scanning with a line sensor or by carrying a two-dimensional sensor.

[0085] In this embodiment, a case where the object to be measured is a micro LED wafer in which a plurality of LED elements each having a size of several tens of μm are formed on a substrate having a diameter of approximately 100 mm is exemplified.

[0086] The data processing device 3 includes a CPU 31 , a RAM 32 , a storage unit 33 , an HSI camera control unit 34 , a data receiving unit 35 , a defect detection unit 36 ​​, a measurement data file generation unit 37 , and a measurement data analysis unit 38 .

[0087] The CPU 31 is a processor that centrally controls the entire data processing device 3 by operating according to an operating program stored in the storage unit 33 and loaded into the RAM 32 .

[0088] The RAM 32 is a memory that provides a work area when the CPU 31 operates according to an operation program.

[0089] The storage unit 33 stores an operation program of the CPU 31 , various application programs, data, and the like, and is composed of an SSD (Solid State Drive), an HDD (Hard Disk Drive), and the like.

[0090] The HSI camera control unit 34 controls the operation of the HSI camera 2 from the start to the end of measurement.

[0091] The data receiving unit 35 receives the measurement data measured by the HSI camera 2 and transmitted from the HSI camera 2 under the control of the CPU 31 .

[0092] The defect detection unit 36 ​​detects the occurrence of data defect in the measurement data received by the data reception unit 35. The detection method will be described later.

[0093] The measurement data file generation unit 37 generates a measurement data file for the measurement data of the HSI camera 2 received by the data receiving unit 35. The measurement data file is generated by aggregating the measurement data for each predetermined unit, for example, each predetermined area. When the measurement data file is generated, when the defect detection unit 36 ​​detects a defect in the data, a specific value that can identify the defective data is embedded in the defect generation location by the method described later. The generated measurement data file is stored in the storage unit 33.

[0094] The measurement data analysis unit 38 reads the measurement data file generated by the measurement data file generation unit 37 and stored in the storage unit 33 from the storage unit 33, and performs various analysis processes as post-processing. As specific examples of the analysis process, it is possible to cite the image production process of the LED element of the micro LED wafer as the measurement object and the analysis process of the spectral characteristics. The analysis process will be described later.

[0095] In addition, the functions of the HSI camera control unit 34 , the defect detection unit 36 ​​, the measurement data file generation unit 37 , and the measurement data analysis unit 38 are realized by the CPU 31 operating according to the operating program.

[0096] Next, the occurrence of a defect in the measurement data, the detection of the data defect by the defect detection unit 36 ​​, and the generation of the measurement data file by the measurement data file generation unit 37 will be described.

[0097] [About the occurrence of data loss]

[0098] As described in the above-mentioned background technology, the measurement data of the HSI camera 2 is a considerable amount of data compared to the data obtained by a conventional RGB 3-channel camera. In particular, the amount of data becomes even greater in the case of a measurement object such as a micro LED wafer having a plurality of LED elements formed on the substrate surface.

[0099] Therefore, it is necessary to perform measurements at a high resolution at high speed. However, depending on the specifications of the hardware on the data processing device 3 side, specifically the performance of the data receiving unit 35, the storage unit 33, etc., sometimes the storage of the measured data cannot keep up with the measurement, resulting in data loss in which the measured data is omitted.

[0100] [About data loss detection]

[0101] Figure 2This is a diagram schematically showing a portion of an image of a measurement object using pixels. The rectangular area is the smallest unit of an image captured by the HSI camera 2, namely, a pixel 21. The number of pixels in a vertical row represents the area captured once, and the spectral data of each wavelength is measured in each pixel 21. When data loss occurs, the spectral data of all pixels 21 in the row are lost. Figure 2 As shown by the arrow in the middle, the HSI camera 2 is moved while at least one of the HSI camera 2 and the object to be measured is photographed in a scanning direction orthogonal to the columns of pixels 21. Each time the HSI camera 2 photographs, a serial number is added to the measurement data and the data is managed according to the serial number. Figure 2 In the example, a serial number of n, n+1, n+2, ..., n+6 is assigned each time a shot is taken.

[0102] If data is missing, the corresponding sequence number will be omitted. Figure 2 In the example of , it means that the sequence number n+4 does not exist and the data of the sequence number n+4 is missing. Therefore, the defect detection unit 36 ​​can detect the occurrence of data loss by confirming the presence or absence of the sequence number added to the measurement data received by the data receiving unit 35, and can determine the location where the data is missing.

[0103] exist Figure 3 The measurement results when data loss occurs are shown in FIG. Figure 3 (A) shows a part of a micro LED wafer 4 as a measurement object, in which a plurality of LED elements 42 are produced on a rectangular substrate 41 . Figure 3 (B) schematically shows an image corresponding to one LED element 42 and its surroundings using a plurality of pixels 21, where each rectangular area represents a pixel. In addition, an area 5 surrounded by a thick frame inside the shaded area corresponds to one LED element 42. In the case of a data loss of one line, as shown in FIG. Figure 3 As shown in (C), it becomes Figure 3 Compared with the original data size shown in (B), the data is shrunk by one line in the scanning direction. If the data is saved in this state, the measurement result will be different from the original data size.

[0104] Therefore, if Figure 3 As shown in (D), the measurement data file generation unit interpolates data by filling the data of all pixels in the row where data loss occurs with a certain value to make the data size consistent with the original data size. The determined value used may be any value as long as it can be distinguished from the actual measured value and can identify the missing data. Through this determined value, the location where the data loss occurs can be identified in the subsequent analysis processing performed by the measurement data analysis unit 38. Figure 3In (D), an example of embedding a digital "0" as a representative of a determined value that cannot be an actual measured value is shown. The determined value also depends on the constructed data processing system 1. Therefore, if the maximum value, minimum value, etc. cannot be an actual measured value in the system 1, using this value, even if the location where the data defect occurs is not separately managed, the location where the data defect occurs can be detected and determined from the measurement data.

[0105] Furthermore, the determination values ​​do not need to be all shared in one determination data file, and different types of determination values ​​may be mixed.

[0106] [Analysis Processing]

[0107] Next, an analysis process will be described as an example of post-processing performed by the measurement data analysis unit 38. This process is performed using a measurement data file generated by the measurement data file generation unit 37 by embedding a certain value at a location where data loss occurs and stored in the storage unit 33.

[0108] Regarding the analysis of the micro LED wafer 4, there are cases where an RGB image or a 1-channel grayscale image is generated and visually confirmed, and cases where the spectral data is analyzed in units of each LED element 42 or in units of a certain data area. Here, the operation for the location where data defects occur in each analysis process is shown.

[0109] 1. When the image is generated

[0110] When generating an image, it is sometimes necessary to clearly indicate the location where data loss occurs. This is necessary, for example, when you want to confirm the location and condition of the data loss. In this case, in order to make it easier to understand the location where the data loss occurs, it is displayed using a numerical value that is completely different from the surrounding area.

[0111] exist Figure 4 This example is shown in (A)(B). Figure 4 (A) and Figure 3 (B) Similarly, an image corresponding to one LED element 42 and its surroundings is schematically represented by a plurality of pixels 21. Each rectangular region represents a pixel. In addition, a region 5 surrounded by a thick frame inside a shaded region corresponds to one LED element 42.

[0112] exist Figure 4 In (A), data loss occurs in one row, and "0" is embedded in the pixel at the location where the data loss occurs. In addition, grayscale spectral data is displayed for pixels that are not lost, in other words, a grayscale image is displayed.

[0113] Figure 4 (B) is for Figure 4The grayscale image (A) shows the location where the data defect occurs at the maximum value (white). By changing the color of the location where the data defect occurs in this way, the location where the data defect occurs can be clearly confirmed. In addition, in the case of an RGB image, a color that does not exist there can also be used, such as green for a blue wafer.

[0114] As a case where it is not necessary to explicitly indicate the location where the data defect occurs, there is a case where the light emitting area of ​​the LED element 42 is specified. In this case, if the location where the data defect occurs is explicitly indicated as described above, there is a possibility that erroneous detection occurs in the detection process of the light emitting area of ​​the LED element 42. In addition, even though it is one LED element 42, it may be determined to be separated.

[0115] Therefore, in such a case, Figure 4 As shown in (C), interpolation based on averaging is performed using the values ​​before and after the generated defective data, and the pixel value is adopted. Figure 4 As shown in (D), image data without data loss can be generated. Therefore, the influence on the detection process of the light emitting area of ​​the LED element 42 can be suppressed to the minimum.

[0116] Furthermore, the interpolation method may be at least one of linear interpolation, spline interpolation, and Lagrange interpolation. In addition, when the location where the data defect occurs is the end of the region, the data before or after the location may be directly used as the pixel value.

[0117] 2. Spectroscopic data analysis

[0118] When analyzing spectroscopic data, it is preferable not to use data of a location where data loss occurs. Therefore, when a location where data loss occurs is detected during analysis, the determined value is excluded from the analysis object and is not included in the numerical calculation.

[0119] For example Figure 5 As shown in (A), when data loss of one line occurs, the pixels corresponding to the LED element 42 to be analyzed (the pixels in the bold area 5 in the figure) where the data loss occurs are excluded. If the pixel values ​​of other pixels without data loss are used, the LED element 42 can be analyzed.

[0120] However, if data loss occurs continuously, there may be cases where the analysis cannot be performed. Figure 5 As shown in (B), if continuous data loss occurs that is larger than the size (in this example, the width) of one LED element 42 in the scanning direction, it is completely impossible to analyze all the LED elements 42 existing in the row.

[0121] Therefore, it can be configured as follows: during analysis, the number of pixels in the analysis object area without data loss is counted, and when the number of pixels is less than the specified number N, it is set as unresolvable, and a warning message is added to the analysis result. The specified number N depends on the accuracy of the data processing system. If sufficient measurement accuracy can be obtained even if the number of pixels in the analysis object area is 1, then N=1 is set. If the accuracy cannot be maintained unless it is above a certain number of pixels, the number of pixels is set to N. In addition, the number of pixels in the analysis object area without data loss is counted, and when the number of pixels is less than the specified number N, it is set as unresolvable. However, it can also be configured as follows: the number of pixels with data loss is counted, and when the number of pixels is greater than the specified number, it is set as unresolvable, and a warning message is added to the analysis result.

[0122] In addition, when the HSI camera 2 measures the spectroscopic data, the number of rows where data defects occur (the number of occurrences) can be counted. Therefore, the defect detection unit 6 can count the number of occurrences of data defects while measuring to determine whether data defects exceeding a specified value occur continuously. It is also possible to report an error when data defects exceeding a specified value occur continuously. The continuous occurrence of data defects is an unresolvable area, but the number of continuous occurrences of defects when the unresolvable area exceeds the size of the LED element can be pre-set to a specified value.

[0123] When data loss of a predetermined value or more occurs continuously, the HSI camera control unit 34 may immediately interrupt the measurement of the HSI camera 2 to check the status, or control the HSI camera 2 to re-measure the measurement portion including the area with continuous loss in real time.

[0124] In addition, in the method of determining whether data defects exceeding a predetermined value are continuously generated at the same time as the measurement, it is impossible to determine whether the continuous data defects are exactly corresponding to the portion where the LED element 42 is present. However, the interval between the LED elements 42 is generally narrower than the size of the LED element 42, so when data defects are continuously generated, there is a high probability that the unresolvable region extends to the portion of the LED element 42. Therefore, determining whether data defects exceeding a predetermined value are continuously generated during measurement is effective in saving the trouble of re-measuring after the measurement is completed.

[0125] Figure 6 1 is a flowchart showing a process of detecting data loss and generating a measurement data file executed by the data processing device 3 . Figure 7The processes shown in these flowcharts are executed by the CPU 31 of the data processing device 3 operating according to an operating program stored in the storage unit 33 or the like and loaded into the RAM 32 .

[0126] exist Figure 6 In step S01, it is determined whether measurement data is received from the HSI camera 2. Measurement data is sent each time one line is measured. If measurement data is not received (No in step S01), it stays in step S01 and waits until measurement data is received.

[0127] If the measurement data is received (yes in step S01), then in step S02, it is checked whether the serial number attached to the measurement data increases continuously, in other words, whether data loss occurs. If the serial number increases continuously (yes in step S02), no data loss occurs. Therefore, in step S03, after directly storing in the storage unit 33 and generating a measurement data file, the process proceeds to step S04.

[0128] In step S02, if the serial number does not increase continuously (No in step S02), data loss occurs. Therefore, in step S05, it is determined whether the number of consecutive occurrences of data loss (the number of consecutive rows of data loss) is greater than a specified value. If it is greater than a specified value (Yes in step S05), it is determined that the continuous data loss area (unresolvable area) is greater than the size of the LED element 42. In step S06, since there is a possibility of a resolution error, after instructing the HSI camera 2 to control the measurement unit including the area where the data loss is continuous, it returns to step S01. It is also possible to instruct the interruption of the measurement without instructing the re-measurement.

[0129] In step S05, if the number of consecutive occurrences of data loss is not greater than the specified value (No in step S05), a determination value is embedded in the location where the data loss occurs in step S07. Then, the data is stored in the storage unit 33 in step S03, and the process proceeds to step S04.

[0130] In step S04, it is checked whether the measurement of the HSI camera is completed. If not completed (No in step S04), the process returns to step S01. If completed (Yes in step S04), the process ends and transfers to step S02. Figure 7 's analysis process.

[0131] exist Figure 7In the analysis process of , it is determined in step S21 whether the image generation process is implemented. If the image generation process is implemented (yes in step S21), the image generation is started in step S22. Then, in step S23, one line of measurement data is read from the measurement data file stored in the storage unit 33 and expanded in the RAM 32. And, in step S24, it is determined whether there is a determined value in the read data. If there is no determined value (no in step S24), the process proceeds to step S26. If there is a determined value (yes in step S24), in step S25, after the data defect occurrence location where the determined value exists is stored, the process proceeds to step S26.

[0132] After the data is directly converted into pixel values ​​in step S26, it is checked in step S27 whether the image generation of all data is completed. If it is not completed (No in step S27), it returns to step S23 and reads the next 1 line of measurement data. If the image generation of all data is completed (Yes in step S27), it proceeds to step S28.

[0133] In step S28, it is checked whether the location where the data defect occurs is explicitly indicated. In the case of explicit indication (yes in step S28), in step S29, the location where the data defect occurs is maintained as a certain value, or converted into a specified fixed value and embedded. For example, in the case of a grayscale image, the data defect portion is explicitly indicated in white. Then, the process proceeds to step S31.

[0134] In step S28, if the location where the data loss occurs is not indicated (No in step S28), in step S30, the pixel values ​​of the data area before and after the use are interpolated, etc., for the location where the data loss occurs. Then, in step S31, the image generation is completed and the analysis process ends.

[0135] On the other hand, if the image generation process is not performed in step S21 (No in step S21), the analysis of the measurement data is started in step S32. Next, in step S33, after the measurement data of the analysis object is read, in step S34, it is determined whether there is a certain value in the read data. If there is no certain value (No in step S34), after the analysis of the target area is performed in step S38, the process proceeds to step S39. If there is a certain value (Yes in step S34), in step S35, the pixels of the certain value are set as outside the analysis object, and in step S36, it is determined whether the number of pixels of the analysis object is less than the specified value N.

[0136] If it is not less than the specified value N (No in step S36), the object area is analyzed in step S38, and then the process proceeds to step S39. If it is less than the specified value N (Yes in step S36), the object area is set as an analysis error in step S37, a warning message is added to the analysis result, and then the process proceeds to step S39.

[0137] In step S39, it is checked whether the analysis processing of all the data has been completed. If it has not been completed (No in step S39), the process returns to step S33 to read the measurement data of the next analysis target.

[0138] If the analysis process of all the data is completed (YES in step S39 ), the analysis process of the measurement data is completed and ends in step S40 .

[0139] Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment. For example, each embodiment may not be independent of each other, and a structure combining two or more embodiments may be possible.

[0140] According to the present embodiment, a determination value capable of identifying the missing data is embedded in the data missing portion of the measurement data file generated by the measurement data file generation unit 37. Therefore, when the measurement data file is used for subsequent processing such as analysis processing, the measurement data analysis unit 38 can identify the situation where the measurement data is missing. Therefore, the interpolation processing in the subsequent processing can of course perform various operations such as indicating the location where the data is missing or ignoring the data. Therefore, appropriate processing corresponding to the content of the subsequent processing can be performed.

[0141] This application claims priority based on Japanese patent application No. 2022-089673 filed on June 1, 2022, the disclosure of which directly constitutes a part of this application.

[0142] Industrial Applicability

[0143] The present invention can be used for a data processing device or the like that generates a measurement data file based on measurement data of a hyperspectral device that can measure two-dimensional spectroscopic data of a measurement object.

[0144] Explanation of the reference numerals: 1…data processing system; 2…HSI camera (HSI device); 21…pixel; 3…data processing device; 31…CPU; 32…RAM; 33…storage unit; 34…HSI camera control unit; 35…data receiving unit; 36…defect detection unit; 37…measurement data file generation unit; 38…measurement data analysis unit; 4…micro LED wafer (measurement object); 41…substrate; 42…LED element; 5…area corresponding to the LED element.

Claims

1. A data processing device, wherein: have: A receiving unit receives measurement data from a hyperspectral device capable of measuring two-dimensional spectroscopic data of a measurement object; a detection unit, configured to detect generation of defects in the measurement data received by the receiving unit; as well as The file generating unit generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a location where the defect has occurred, when the occurrence of defect in the measurement data is detected by the detecting unit.

2. The data processing device according to claim 1, wherein: An analyzing unit is provided, the analyzing unit analyzes the measurement data file generated by the file generating unit, The analyzing unit switches the operation of the defect occurrence location of the measurement data file according to the purpose of analysis.

3. The data processing device according to claim 1 or 2, wherein: The determined value is 0 or a maximum value or a minimum value that can be held with respect to the numerical value of the measurement data.

4. The data processing device according to claim 2, wherein: The analysis unit interpolates the measurement data at the defect generation location by using at least any one interpolation method among linear interpolation, spline interpolation and Lagrange interpolation.

5. The data processing device according to claim 2, wherein: The detection unit counts the occurrence of defects and determines whether a predetermined value or more of defects set in advance has been continuously generated.

6. The data processing device according to claim 5, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, and the predetermined value is the number of times defects are generated when an unresolvable area generated due to continuous generation of defects exceeds the size of the LED elements.

7. The data processing device according to claim 5, wherein: A control unit is provided. When the detection unit determines that defects exceeding a preset specified value are continuously generated, the control unit controls the hyperspectral device to interrupt the measurement performed by the hyperspectral device or re-measure the measurement section including the area where the defects are continuous.

8. The data processing device according to claim 2, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, The analysis unit generates an image of the micro LED wafer based on the measured spectral data, and determines that data is missing when the determined value is detected during image generation, and directly uses the determined value or replaces it with other values ​​or uses previous and subsequent data for interpolation.

9. The data processing device according to claim 2, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, The analyzing unit performs analysis in units of LED elements on the micro LED wafer, and when the specific value is detected, excludes pixels of the specific value from analysis targets.

10. The data processing device according to claim 9, wherein: The analysis unit adds warning information to the analysis result of the LED element when the number of pixels not excluded from the analysis target is less than a predetermined value.

11. A data processing system, wherein: A hyperspectral device capable of measuring two-dimensional spectral data of a measurement object, and the data processing device according to claim 1 or 2.

12. A data processing method, wherein: include: A measuring step, measuring two-dimensional spectroscopic data of a measuring object by a hyperspectral device; a detection step of detecting generation of defects in the measurement data measured by the measurement step; as well as The file generating step generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a location where the defect has occurred, when the occurrence of defect in the measurement data is detected in the detecting step.

13. A program, wherein: Used to cause a computer to: A receiving step of receiving measurement data from a hyperspectral device capable of measuring two-dimensional spectroscopic data of a measurement object; a detecting step of detecting generation of defects in the measurement data received in the receiving step; as well as The file generating step generates a measurement data file in which a specific value capable of identifying the defective data is embedded in a location where the defect has occurred, when the occurrence of defect in the measurement data is detected in the detecting step.

14. The program according to claim 13, wherein: further causing the computer to execute an analysis step in which the measurement data file generated by the file generation step is analyzed, In the analysis step, the computer is caused to execute a process of switching an operation of a location where a defect of the measurement data file occurs according to an analysis purpose.

15. A program according to claim 13 or 14, wherein: The determined value is 0 or a maximum value or a minimum value that can be held with respect to the numerical value of the measurement data.

16. The program according to claim 13 or 14, wherein: In the analysis step, the computer is caused to execute the following processing: at the defect generation site, the measurement data is interpolated by at least one interpolation method among linear interpolation, spline interpolation and Lagrange interpolation.

17. The program according to claim 14, wherein: In the detection step, the computer is caused to execute a process of counting the occurrence of defects and determining whether a number of defects exceeding a predetermined value set in advance has been continuously generated.

18. The program according to claim 17, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, and the predetermined value is the number of times defects are generated when an unresolvable area generated due to continuous generation of defects exceeds the size of the LED elements.

19. The program according to claim 17, wherein: The computer is caused to execute a control step. In the control step, when it is determined by the detection step that defects exceeding a preset specified value are continuously generated, the hyperspectral device is controlled so as to interrupt the measurement performed by the hyperspectral device or re-measure the measurement portion including the area where the defects are continuous.

20. The program according to claim 14, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, In the analysis step, the computer is caused to perform the following processing: image data of the micro LED wafer is generated based on the measured spectral data, and when the determined value is detected during image generation, it is determined that the data is missing, and the determined value is directly used or replaced with other values ​​or interpolated using the previous and next data.

21. The program according to claim 14, wherein: The object to be measured is a micro LED wafer having a plurality of LED elements, In the analysis step, the computer is caused to execute the following processing: analyzing in units of LED elements on the micro LED wafer, and when the specific value is detected, excluding pixels of the specific value from the analysis target.

22. The program according to claim 21, wherein: In the analysis step, the computer is caused to execute a process of adding warning information to the analysis result of the LED element when the number of pixels not excluded from the analysis target is smaller than a predetermined value.

Images