Camera debugging dump original data classification method and device

Abstract

The invention provides a camera debugging dump original data classification method and device, and the method comprises the steps: classifying the dump data of each scene into a respective folder, carrying out the recognition and correspondence of a jpg picture of a dump file in the folder and a jpg picture with a serial number, extracting the debugging information of the jpg, and displaying the debugging information in the name of the folder. The file folder named by the serial number of each scene and the debugging information is obtained, when a certain scene needs to be debugged, the file folder corresponding to the scene can be directly found according to needs, and a debugger can debug the corresponding scene conveniently. The problems that the number of mixed dump files of all cameras is large, scene dump files are difficult to find and similar scenes cannot be associated in the current mobile phone camera debugging process can be solved.

Description

Technical Field

[0001] The invention belongs to the field of mobile phone camera debugging, relates to image processing, hardware testing and other aspects, and in particular to a camera debugging dump raw data classification method and device. Background Art

[0002] With the rapid development of technology, mobile phone cameras have become an indispensable part of our daily lives, with a wide range of applications, from simple photo taking to complex image processing and analysis. In order to meet users' demand for high-quality images, a large number of subjective scene tests are required in the research and development phase of mobile phone cameras to ensure that all possible user scenarios are covered. In the current mobile phone camera image quality debugging process, debuggers need to obtain the dump data of a single scene and import it into the simulation tool for debugging. The dump data contains a variety of data such as jpg, raw, awb, lsc, gamma, color, etc. The file names of these data contain specific strings and are named with timestamps and file names.

[0003] When the dump command is turned on to take a photo, the parameters of each module of the ISP (image signal processor) will be generated. These parameters are saved in a separate file for simulation. The debugger needs to remember the serial number of the problem scene reported by the test end, then find the corresponding jpg file in the dump file package, copy its anonymized name, find the corresponding raw (original) image and import it into the simulation tool for debugging. However, this method has some disadvantages: first, due to the large number of mixed dump files of each camera, it is difficult for debuggers to quickly find the dump data of a specific scene; second, the data of multiple similar scenes cannot be effectively associated, resulting in low debugging efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a camera debugging dump raw data classification method and device, aiming to solve the problems of a large number of mixed dump files in each camera, difficulty in finding scene dump files, and inability to associate similar scenes during the current mobile phone camera debugging process.

[0005] To achieve the above object, on the one hand, the present invention provides a camera debugging dump raw data classification method, the method comprising the steps of:

[0006] Export the dumped raw data generated by shooting multiple scenes;

[0007] All dumped raw data are sorted into individual folders by scene based on timestamp and file name characteristics;

[0008] Extract the debugging information required for debugging, add it to the name of the folder corresponding to the single scene, and place the dump file of the single scene in the folder;

[0009] Arrange the jpg images of each scene in sequence;

[0010] Compare and identify the jpg images of the dump files in the folder with the jpg images of the scenes, match serial numbers for each scene, and add the matched serial numbers to the folder names corresponding to each scene.

[0011] Furthermore, all the dumped raw data are classified into a single folder by scene, including:

[0012] Extract files ending with .jpg from all dump files;

[0013] A timestamp in a specific format is intercepted from the dump file name as the first valid bit of timestamp information of the jpg file, and debugging information is extracted from the jpg file.

[0014] Furthermore, the debugging information includes at least one of CCT color temperature, BV brightness, ISO information and picture size.

[0015] Furthermore, the step of classifying all the dumped raw data into a single folder according to the scene also includes:

[0016] Create a jpg information dictionary and output the jpg information dictionary.

[0017] Furthermore, the key of the jpg information dictionary is the first valid bit timestamp of each jpg file, and the value is the corresponding debugging information.

[0018] Furthermore, the classifying all the dumped raw data into a single folder according to the scene also includes traversing all the dumped data, and when a single dumped data is input, intercepting a timestamp in a specific format from the dumped file name as the second significant bit timestamp information of the dumped data.

[0019] Further, extract the keys in the jpg information dictionary to create a jpg timestamp list, and check whether the second valid timestamp information exists in the jpg timestamp list;

[0020] If it exists, create a new folder name according to the first valid digit timestamp + the value corresponding to the first valid digit timestamp of the dump file in the jpg information dictionary, and move the corresponding dump file to the folder;

[0021] If not present, the dump file is discarded.

[0022] Furthermore, the method of classifying all the dumped raw data into a single folder according to the scene also includes classifying the folders according to the image size, corresponding to the debugging requirements of different cameras such as front camera, rear camera, wide angle, macro, and depth of field.

[0023] Furthermore, the jpg image of the dump file in the folder is compared and identified with the scene jpg image to accurately determine whether the scene jpg image and the jpg image in the folder belong to the same scene, and the steps are as follows:

[0024] Input scene jpg images arranged by sequence number and extract the DateTime value of each scene jpg image;

[0025] Create a dictionary of jpg serial numbers and DateTime values;

[0026] Traversing the dump files of each scene that has been classified into the folder, and extracting the first valid bit timestamp of the folder;

[0027] Subtract the first valid timestamp of the scene from the DateTime value of all jpg images;

[0028] Determine whether the difference is less than the set threshold to determine whether the jpg image and the dump file may belong to the same scene;

[0029] If the difference is less than the threshold, the jpg image and the jpg image in the dump file are subtracted through image comparison technology to determine whether they are the same scene;

[0030] If the difference between two images is 0, then the two images are images of the same scene.

[0031] Furthermore, in the step of traversing the dump files of various scenes that have been classified into the folder, the first valid bit timestamp of the folder is the time when the dump file is generated.

[0032] Furthermore, when the two images are images of the same scene, the serial number information of the jpg images is added to the corresponding folder.

[0033] In another aspect, the present invention provides a data classification device, the device comprising:

[0034] Classification module, used to classify the dump files of each scene into their respective folders;

[0035] The corresponding recognition module is used to identify and correspond the classified folders with the numbered jpg images;

[0036] The debugging information extraction module is used to extract the debugging information of the jpg image and display it in the folder name;

[0037] Data moving module, used to move dump files to corresponding folders;

[0038] The timestamp matching module is used to extract the timestamp information of jpg images and dump files and match them;

[0039] The image comparison module is used to compare the jpg image with the jpg image in the dump file using image processing technology to confirm whether they belong to the same scene;

[0040] The scene optimization module is used to find dump data of similar scenes based on the extracted debugging information to facilitate parameter conflict optimization.

[0041] Compared with the prior art, the present invention has at least the following technical effects:

[0042] The present invention provides a camera debugging dump raw data classification method and device, which classifies the data in the dump file of each scene into respective folders, identifies and corresponds the jpg images of the dump file in the folder with the jpg images with arranged serial numbers, and extracts the jpg debugging information to display in the folder naming, so as to obtain a folder named with the serial number of each scene. When a certain scene needs to be debugged, the folder of the scene can be directly found as needed, which is convenient for the debugger to debug the corresponding scene. Thereby, the data in the dump file can be accurately corresponded to the scene, and the scene can be classified, so that the debugger can quickly find and optimize the problem scene. In addition, through the automated classification method, the present invention can quickly associate the dump data with the corresponding scene jpg, thereby greatly reducing the time spent by the debugger in finding and matching data, thereby improving the overall work efficiency.

[0043] Furthermore, by creating a jpg information dictionary, it is possible to quickly index and retrieve jpg information, providing efficient data structure support for dump file data classification and scene matching, and facilitating subsequent association and analysis of similar scenes. In addition, using image comparison technology to quickly determine whether the scene jpg image and the jpg image in the dump file of the folder are the same scene can greatly improve the speed and accuracy of scene matching, reduce human judgment errors, and provide efficient technical support for the processing of large-scale dump file data. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A detailed flow chart of a camera debugging dump raw data classification method according to an embodiment of the present invention;

[0045] Figure 2A general flow chart of a camera debugging dump raw data classification method in one embodiment of the present invention;

[0046] Figure 3 The file directory structure generated during the camera ISP debugging process in one embodiment of the present invention;

[0047] Figure 4 A flowchart of classifying the dumped originals into various folders according to scenes in one embodiment of the present invention;

[0048] Figure 5 The present invention is a flowchart of a method for determining whether a scene jpg and a jpg in a folder dump file are the same in one embodiment of the present invention. DETAILED DESCRIPTION

[0049] A camera debugging dump raw data classification method and device of the present invention will be described in more detail below in conjunction with a schematic diagram, wherein a preferred embodiment of the present invention is shown, and it should be understood that those skilled in the art can modify the present invention described herein while still achieving the advantageous effects of the present invention. Therefore, the following description should be understood as being widely known to those skilled in the art, and not as a limitation of the present invention.

[0050] For the sake of clarity, not all features of the actual embodiments are described. In the following description, well-known functions and structures are not described in detail because they would clutter the invention with unnecessary detail. It should be recognized that in the development of any actual embodiment, a large number of implementation details must be made to achieve the developer's specific goals, such as changing from one embodiment to another according to the limitations of the relevant system or the relevant business. In addition, it should be recognized that such development work may be complex and time-consuming, but it is just a routine task for those skilled in the art.

[0051] The present invention is described in more detail in the following paragraphs by way of example with reference to the accompanying drawings. The advantages and features of the present invention will become more apparent from the following description. It should be noted that the accompanying drawings are in very simplified form and are not in exact proportions, and are only used to facilitate and clearly assist in illustrating the purpose of the embodiments of the present invention.

[0052] As mentioned in the background technology, the current mobile phone camera debugging process has problems such as a large number of mixed dump files for each camera, difficulty in finding scene dump files, and inability to associate similar scenes. In view of this, on the one hand, Figure 1 As shown, the present invention provides a camera debugging dump raw data classification method to solve the above problem. Specifically, the method comprises the following steps:

[0053] Export the dumped raw data generated by shooting multiple scenes;

[0054] All dumped raw data are sorted into individual folders by scene based on timestamp and file name characteristics;

[0055] Extract the debugging information required for debugging, add it to the name of the folder corresponding to the single scene, and place the dump file of the single scene in the folder;

[0056] Arrange the jpg images of each scene in sequence;

[0057] Compare and identify the jpg images of the dump files in the folder with the jpg images of the scenes, match serial numbers for each scene, and add the matched serial numbers to the folder names corresponding to each scene.

[0058] Furthermore, this embodiment provides a general process of classifying camera debugging dump data as follows: Figure 2 As shown, (it should be noted that, in this embodiment, the dump instruction represents a dump instruction, and dump represents dump data) the specific steps are as follows:

[0059] Step 001: A mobile phone that captures the entire scene

[0060] The debugging machine opens the dump command (dump command). The dump command is used to open the saving function of some ISP process files. When taking a photo, some parameters of the ISP modules called when taking this photo will be generated. These parameters will be saved in a separate file for simulation. When raw is selected during simulation, these files will be imported into the simulation tool together. Generally, these files will be generated in a specific file, such as the directory of the mobile phone: data / vendor / camera_dump. The dump data generated by a single scene contains: jpg, raw, awb, lsc, gamma, color and other data, such as Figure 3 When shooting multiple scenes, export data / vendor / camera_dump, which is called dumping raw data here, and may also be referred to as dumping raw data below.

[0061] Step 002: Sort all dump data into a single folder

[0062] When shooting according to a single scene, the dumped data will be named with timestamp + file name. This feature will be used to move the dumped original data of the same scene to a single folder later.

[0063] Step 003: Name the folder:

[0064] Extract the debugging information required for debugging and add it to the folder naming, such as CCT, BV, ISO, Size, etc.

[0065] Step 004: Output the folder with information naming

[0066] Place all dump files of a single scene in the corresponding folder named after its information. The folder name contains: the timestamp information of the generation of these dump files + CCT_xxx_BV_xxx_ISO_xxx_Size_xxxx, etc.

[0067] Step 005: Numbered scene jpg images

[0068] Number the full-scene pictures of test machine 1, test machine 2, test machine 3, comparison machine 1, and comparison machine 2 for front camera, rear camera, wide angle, macro, and depth of field according to the scene, and use this number to evaluate, check for picture problems, and record scene problems.

[0069] Step 006: Output the dump data packet corresponding to each scene

[0070] By comparing and identifying the dumped jpg images in the folder with the jpg, a matching serial number is assigned to each scene, and the matching serial number is recorded in the folder name corresponding to each scene.

[0071] In this embodiment, after executing the above steps 001-006, each scene will get a folder named with information, and the dump data of each scene (including jpg, raw, awb, lsc, gamma, color, etc.) will be classified and stored in the folder named with information. The information includes CCT (Correlated Color Temperature) color temperature information, BV (Brightness Value) brightness information, ISO (sensitivity) information and Size (size) information.

[0072] In this embodiment, all dumped raw data are classified into a single folder by scene, which can improve data management efficiency, reduce the time for debuggers to find dump data for a specific scene, facilitate rapid identification and association of similar scenes, facilitate parameter optimization and troubleshooting, and intuitively display key debugging information through folder naming, thereby improving debugging efficiency.

[0073] Specifically, when a scene needs to be debugged, the folder of the scene can be directly found as needed, which is convenient for the debugger to debug the corresponding scene. This means that AWB (white balance) debuggers can find scenes with similar color temperature and brightness for simulation by viewing the CCT color temperature and BV brightness information in the folder naming. The advantage of this is that similar lighting conditions and brightness levels may have similar white balance requirements, so you can focus on optimizing the AWB performance in these scenes. ISP debuggers can find scenes with similar ISO values ​​for simulation by viewing the ISO information in the folder naming. This helps to focus on optimizing image quality at specific ISO settings, such as noise control, dynamic range, etc.

[0074] Furthermore, since the dump data generation rule is that after taking a picture, a jpg, raw and other ISP module data will be generated in the dump path, this embodiment uses whether the jpg in the dump path is the same as the jpg of each scene as the judgment basis. If they are the same, it means that this dump is the dump of the scene.

[0075] Specifically, the detailed implementation method of step 002 is as follows Figure 4 As shown, the specific steps are as follows:

[0076] Step 101: Dump all raw data

[0077] After the test machine is tested, the dump in its dump path is exported. The dump contains a large amount of dump data of each module test.

[0078] Step 102: Extract jpg data from dump raw

[0079] Due to some performance or lag factors, the dump generated by the shooting scene is sometimes incomplete or disordered, so this embodiment adopts a method of extracting jpg data from all dump files, which uses a for loop to determine the end of all dump files. When it ends with .jpg, the file is extracted.

[0080] Step 103: Extract valid bit timestamp information

[0081] In the dump data, the original jpg name format is test time + serial number, for example: 173739035-0283-0000-173743584.jpg. After checking the dump files of the same scene, it is found that xxxxx-xx is the same. Therefore, the second "-" in the dump file is used for string interception, and the data on its left is selected as the first valid bit timestamp of the image. At the same time, the exif (debugging information) of the jpg is extracted. The extracted debugging information includes CCT color temperature information, BV brightness information, ISO information, and image size.

[0082] Step 104: Create a jpg information dictionary

[0083] The debugging information corresponding to the jpg in step 103 is stored in the jpg information dictionary, wherein the key of the jpg information dictionary is the first valid timestamp of the jpg, and the value is the debugging information extracted from the corresponding jpg file. Therefore, the debugging information corresponding to each jpg can be obtained according to the jpg information dictionary, for example, 173739035-0283-0000-173743584.jpg obtains the dictionary {"173739035-0283":"CCT_xxx_BV_xxx_ISO_xxx_Size_xxxx"}.

[0084] Step 105: Output jpg information dictionary

[0085] Extract all jpg images from all dumps and get the final jpg information dictionary.

[0086] Step 201: ALL dump raw data

[0087] After the test machine is tested, the dump in its dump path is exported. The dump contains a large amount of dump data of each module test.

[0088] Step 202: Single dump data is passed in

[0089] Traverse the data in ALL dump raw data so that a single dump data is passed in.

[0090] Step 203: Extract the second valid bit timestamp information

[0091] The original dump data format is test time + serial number, for example: 100641844-0172-0175-main-Dip_IspVirBuf-TG1-DQ55207-SZ49152-MFNR_MSF_5_p2.reg. When a single dump data is passed in, the second "-" in the dump file is used to intercept the string, and the data on its left is selected as the second significant bit timestamp of the image, which is 100641844-0172.

[0092] Step 204: Determine whether the time stamp is in the timestamp list

[0093] In step 105, all dumped jpg data, that is, the information dictionary of all jpgs, has been obtained, and the key value key is extracted as a jpg timestamp list (list), and it is determined whether the second valid bit timestamp information in 203 is in the jpg timestamp list.

[0094] If it exists, the new folder name will be composed of the first valid timestamp + the value corresponding to the first valid timestamp of the file in the jpg information dictionary. For example: 100641844-0172-0175-main-Dip_IspVirBuf-TG1-DQ55207-SZ49152-MFNR_MSF_5_p2.reg, the new folder name is: 100641844-0172_CCT_xxx_BV_xxx_ISO_xxx_Size_xxxx

[0095] Step 205:

[0096] If the result of the determination in step 204 does not exist, the dump file is discarded, that is, no processing is performed.

[0097] Step 206:

[0098] It is determined whether the folder name generated in step 204 already exists. If it does, proceed to step 208; if it does not, proceed to step 207.

[0099] Step 207:

[0100] If it does not exist, create the folder by establishing the file path.

[0101] Step 208: Move to Folder

[0102] Move the corresponding dump file to the existing folder.

[0103] Step 209: Distinguish to the corresponding size folder

[0104] Move all the distinguished folders to front camera, rear camera, wide angle, macro, depth of field according to size. When debugging each camera, you can copy the camera dump file you want to debug as needed.

[0105] Furthermore, the purpose of step 004 and step 005 is mainly to find out whether the scene jpg image and the jpg image in the folder dump are the same. The method adopted is: use the image comparison technology of OpenCV's subtract function (in OpenCV, the subtract function is used to calculate the element-by-element difference between two images. The purpose of this function is to subtract the corresponding pixel values ​​of one image from another image) to compare the two images. The principle is to perform a pixel-by-pixel subtraction operation of the image, that is, to subtract the two images to obtain the difference. The subtract function needs to read the two images separately. If it is not limited and the function is used to traverse, in the actual subjective test, the number of scene jpgs and jpgs in the folder dump are very large. In mathematics, it is the product of the number of the two. It takes a huge amount of time for the subtract function to read these. Therefore, this embodiment pre-determines the timestamps of the subjective scene jpg and the jpg in the folder dump. When the timestamp difference is within a certain range, the subtract function is used to determine the two, which is fast and accurate. The specific steps are as follows: Figure 5 The detailed steps are as follows:

[0106] Step 301: jpg images of the numbered scenes

[0107] Enter the scene jpg images arranged in sequence, the names of these images have been changed or not.

[0108] Step 302: Extract the DateTime of each jpg

[0109] The exif information of each jpg stores the shooting time of the picture and writes it into exif DateTime. The DateTime value of each picture is read through the function. The DateTime of jpg pictures taken with the same camera is unique.

[0110] Step 303: Create a dictionary of jpg serial numbers and DateTime:

[0111] In step 302, the DateTime of each jpg is read, and the serial number of the jpg and the DateTime are combined into a dictionary, that is, {DateTime: serial number}

[0112] Step 304: Folder named ALL dump with information

[0113] In step 004, the dump files of each scene have been classified into folders, and the folders already contain the first valid timestamp + debugging information.

[0114] Iterate through these folders.

[0115] Step 305: Extract the first valid timestamp information of the folder

[0116] Extract the first valid digit timestamp information of the folder. This first valid digit timestamp information is the time when the dump file of this scene is generated.

[0117] Step 306: Subtract the elements in the DateTime list

[0118] The difference between the first valid bit timestamp of the scene in step 305 and the DateTime of all jpgs with the serial numbers in step 303 is taken. In principle, the time when the jpg is generated by taking a photo and the time when the dump file is generated should be close.

[0119] Step 307: Is the difference less than the set threshold?

[0120] Since the time when the jpg is generated and the time when the dump file is generated should be close, a threshold is set between the time when the jpg is generated and the time when the dump file is generated to determine whether the jpg and the dump are of the same scene. The threshold is fluctuating. The purpose of setting the threshold is to reduce the comparison time. When it is less than the threshold, the jpg and the dump may be of the same scene. When it is greater than the threshold, the jpg is definitely not the dump corresponding to this scene, and the next jpg DateTime is traversed.

[0121] Step 308:

[0122] After step 307, it is determined that the currently traversed jpg may be the same scene as the dump in the folder, and the jpg corresponding to the DateTime and the jpg of the dump file in the folder are extracted respectively.

[0123] Step 309: Determine whether the two photos are of the same scene

[0124] Use OpenCV's subtract function to compare the two images. The principle is to perform pixel-by-pixel subtraction of the image, that is, to subtract the two images and get the difference. If the difference is 0, the two images are images of the same scene, and proceed to the next step. If the difference exists and is not 0, the jpg and dump are not the same scene, and continue to traverse the DateTime of the next jpg.

[0125] Step 310:

[0126] The folder is assigned a serial number. In step 309, it has been determined that the jpg and the dump file in the folder are the same scene, so the jpg serial number information is added to the folder to achieve the correspondence between the scene jpg and the dump file.

[0127] In another aspect of this embodiment, a data classification device is also provided, the device comprising:

[0128] Classification module, used to classify the dump files of each scene into their respective folders;

[0129] The corresponding recognition module is used to identify and correspond the classified folders with the numbered jpg images;

[0130] The debugging information extraction module is used to extract the debugging information of the jpg image and display it in the folder name;

[0131] Data moving module, used to move dump files to corresponding folders;

[0132] The timestamp matching module is used to extract the timestamp information of jpg images and dump files and match them;

[0133] The image comparison module is used to compare the jpg image with the jpg image in the dump file using image processing technology to confirm whether they belong to the same scene;

[0134] The scene optimization module is used to find dump data of similar scenes based on the extracted debugging information to facilitate parameter conflict optimization.

[0135] In summary, in a camera debugging dump raw data classification method and device provided by an embodiment of the present invention, by classifying the data in the dump file of each scene into respective folders, and identifying and corresponding the jpg images of the dump file in the folder with the jpg images with arranged serial numbers, and extracting the jpg debugging information to display in the folder naming, a folder named with the serial number of each scene is obtained. When a certain scene needs to be debugged, the folder of the scene can be directly found as needed, which is convenient for the debugger to debug the corresponding scene. Thereby, the data in the dump file can be accurately corresponded to the scene, and classified by scene, so that the debugger can quickly find and optimize the problem scene. In addition, through the automated classification method, the present invention can quickly associate the dump data with the corresponding scene jpg, thereby greatly reducing the time spent by the debugger in finding and matching data, thereby improving the overall work efficiency.

[0136] Furthermore, by creating a jpg information dictionary, it is possible to quickly index and retrieve jpg information, providing efficient data structure support for dump file data classification and scene matching, and facilitating subsequent association and analysis of similar scenes. In addition, using image comparison technology to quickly determine whether the scene jpg image and the jpg image in the dump file of the folder are the same scene can greatly improve the speed and accuracy of scene matching, reduce human judgment errors, and provide efficient technical support for the processing of large-scale dump file data.

[0137] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A camera debugging dump raw data classification method, characterized in that: The steps of the method include: Export the dumped raw data generated by shooting multiple scenes; All dumped raw data are sorted into individual folders by scene based on timestamp and file name characteristics; Extract the debugging information required for debugging, add it to the name of the folder corresponding to the single scene, and place the dump file of the single scene in the folder; Arrange each scene's jpg pictures in sequence; Compare and identify the jpg images of the dump files in the folder with the jpg images of the scenes, match serial numbers for each scene, and add the matched serial numbers to the folder names corresponding to each scene.

2. The method according to claim 1, characterized in that The above will categorize all dumped raw data into individual folders by scene, including: Extract files ending with .jpg from all dump files; A timestamp in a specific format is intercepted from the dump file name as the first valid bit of timestamp information of the jpg file, and debugging information is extracted from the jpg file.

3. The method according to claim 2, characterized in that The debugging information includes at least one of CCT color temperature, BV brightness, ISO information and picture size.

4. The method according to claim 2, characterized in that: The dumped raw data is sorted into individual folders by scene, and also includes: Create a jpg information dictionary and output the jpg information dictionary.

5. The method according to claim 4, characterized in that The key of the jpg information dictionary is the first valid bit timestamp of each jpg file, and the value is the corresponding debugging information.

6. The method according to claim 5, characterized in that The method of classifying all the dumped raw data into a single folder according to the scene also includes traversing all the dumped data, and when a single dumped data is input, intercepting a timestamp in a specific format from the dumped file name as the second significant bit timestamp information of the dumped data.

7. The method according to claim 6, characterized in that Extract the keys in the jpg information dictionary to create a jpg timestamp list, and check whether the second valid timestamp information exists in the jpg timestamp list; If it exists, create a new folder name according to the first valid digit timestamp + the value corresponding to the first valid digit timestamp of the dump file in the jpg information dictionary, and move the corresponding dump file to the folder; If not present, the dump file is discarded.

8. The method according to claim 1, characterized in that The method further includes classifying all dumped raw data into a single folder according to the scene, and classifying the folders according to the picture size, corresponding to the debugging requirements of different cameras such as front camera, rear camera, wide angle, macro, and depth of field.

9. The method according to any one of claims 2 to 8, characterized in that: The steps of comparing and identifying the jpg image of the dump file in the folder with the scene jpg image to accurately determine whether the scene jpg image and the jpg image in the folder belong to the same scene are as follows: Input scene jpg images arranged by sequence number and extract the DateTime value of each scene jpg image; Create a dictionary of jpg serial numbers and DateTime values; Traversing the dump files of each scene that has been classified into the folder, and extracting the first valid bit timestamp of the folder; Subtract the first valid timestamp of the scene from the DateTime value of all jpg images; Determine whether the difference is less than the set threshold to determine whether the jpg image and the dump file may belong to the same scene; If the difference is less than the threshold, the jpg image and the jpg image in the dump file are subtracted through image comparison technology to determine whether they are the same scene; If the difference between two images is 0, then the two images are images of the same scene.

10. The method according to claim 9, characterized in that In the step of traversing the dump files of the various scenes that have been classified into the folder, the first valid bit timestamp of the folder is the time when the dump file is generated.

11. The method according to claim 9, characterized in that When the two images are images of the same scene, the serial number information of the jpg images is added to the corresponding folder.

12. A data classification device, characterized in that: include: Classification module, used to classify the dump files of each scene into their respective folders; The corresponding recognition module is used to identify and correspond the classified folders with the numbered jpg images; The debugging information extraction module is used to extract the debugging information of the jpg image and display it in the folder name; Data moving module, used to move dump files to corresponding folders; The timestamp matching module is used to extract the timestamp information of jpg images and dump files and match them; The image comparison module is used to compare the jpg image with the jpg image in the dump file using image processing technology to confirm whether they belong to the same scene; The scene optimization module is used to find dump data of similar scenes based on the extracted debugging information to facilitate parameter conflict optimization.

Images