A calibration method for image equalization

Abstract

The present application discloses a calibration method for image equalization, which relates to the field of analysis technology, and includes confirming a target wide-field camera, receiving image data collected by each image sensor corresponding to the target wide-field camera, and combining each image data into a wide-field original image in a preset arrangement, and marking the wide-field original image as a target image; obtaining a reference area corresponding to the target image based on the target image, and confirming the mean of a multi-channel image sensor corresponding to the reference area of ​​the target image, and confirming the row and column correction error coefficient corresponding to the target image according to the mean of the multi-channel image sensor corresponding to the reference area of ​​the target image; correcting the target image according to the row and column correction error coefficient corresponding to the target image, and then extracting the corresponding corrected image after correction. The present application has the effect of improving image quality.

Description

Technical Field

[0001] The present application relates to the field of analysis technology, and in particular to a calibration method for image equalization. Background Art

[0002] Calibration is a key technology in the field of computer vision. By performing image equalization calibration on the images taken by the camera, the computer can more accurately reflect the image information of the actual physical scene.

[0003] In the related art, image algorithms are all aimed at distortion-free and image equalization methods for single-channel image sensor imaging. The processing effects of wide-format photos composed of photos taken by multi-channel image sensors are uneven, and thus high-quality images cannot be provided in image acquisition, and there is room for improvement. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present application provides a calibration method for image equalization.

[0005] In a first aspect, the present application provides an image equalization calibration method, which adopts the following technical solution:

[0006] A calibration method for image equalization comprises the following steps:

[0007] Confirming a target wide-field-of-view camera, receiving image data collected by each image sensor corresponding to the target wide-field-of-view camera, combining each image data into a wide-field-of-view original image in a preset arrangement, and marking the wide-field-of-view original image as a target image;

[0008] Acquire a reference area corresponding to the target image based on the target image, confirm a multi-channel image sensor mean value of the reference area corresponding to the target image, and confirm a row and column correction error coefficient corresponding to the target image according to the multi-channel image sensor mean value of the reference area corresponding to the target image;

[0009] The target image is corrected according to the row and column correction error coefficients corresponding to the target image, and then the corrected image is extracted.

[0010] By adopting the above technical scheme, by obtaining the reference area corresponding to the target image and confirming the multi-channel image sensor mean value of the reference area corresponding to the target image, and confirming the row and column correction error coefficients corresponding to the target image according to the multi-channel image sensor mean value of the reference area corresponding to the target image, the target image is corrected according to the row and column correction error coefficients corresponding to the target image, and then the corrected image is extracted, thereby effectively reducing the situation where high-quality images cannot be collected due to the uneven processing effects of wide-format photos composed of photos taken by multi-channel image sensors.

[0011] Preferably, the preset arrangement is set to an arrangement sequence number corresponding to each image sensor of the target wide-field camera.

[0012] Preferably, obtaining a reference area corresponding to the target image based on the target image specifically includes:

[0013] Get the grayscale histogram corresponding to the target image ,in, Represented as grayscale, , Represented as the gray level in the target image is The number of pixels;

[0014] Analyze the grayscale histogram corresponding to the target image and confirm the preset brightness range Then calculate the uniformity evaluation coefficient corresponding to the preset brightness range , ,in, Indicates the preset brightness range The average number of pixels corresponding to

[0015] Adjust and traverse the preset brightness range, and then confirm the uniform evaluation coefficient corresponding to each traversal, and store the uniform evaluation coefficient corresponding to each traversal in an evaluation set, sort the uniform evaluation coefficients in the evaluation set according to a preset order, confirm the target uniform evaluation coefficient, and then obtain the target brightness range corresponding to the target uniform evaluation coefficient, confirm the corresponding target pixel in the target image based on the target brightness range, and confirm the distribution area corresponding to the target pixel, and then extract the distribution coefficient corresponding to each distribution area, and then confirm the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel.

[0016] Preferably, confirming the distribution area corresponding to the target pixel, and then extracting the distribution coefficient corresponding to each distribution area, and then confirming the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel, specifically includes:

[0017] Obtain the distribution area corresponding to the target pixel and confirm the basic information corresponding to each distribution area, including the number of distribution areas and the number of pixels corresponding to each distribution area. ,in, It is represented by the number corresponding to each distribution area. , Indicated as number The number of pixels corresponding to the distribution area;

[0018] By formula , and the distribution coefficient corresponding to each distribution area is obtained ,in, Represents the number corresponding to each pixel, Expressed as In the distribution area The brightness value corresponding to each pixel is Expressed as The mean brightness of the pixels corresponding to the distribution area;

[0019] Get the distribution coefficient corresponding to each distribution area , the distribution coefficient corresponding to each distribution area With the preset distribution coefficient threshold Compare, if there is Less than , then the distribution area is stored in the reference area set;

[0020] The area corresponding to each distribution area in the reference area set is confirmed, the distribution area with the largest area in the reference area set is obtained, and then the distribution area is set as the reference area corresponding to the target image.

[0021] Preferably, confirming the mean value of the multi-channel image sensor of the reference area corresponding to the target image, and confirming the row and column correction error coefficients corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image, specifically includes:

[0022] Confirm the position information corresponding to the reference area and use express;

[0023] By formula , confirm the multi-channel image sensor mean corresponding to the target image reference area ,in, They are respectively represented as the number of rows and columns corresponding to the pixels in the reference area, Respectively, they are represented as reference areas The positions correspond to the pixel values ​​of the red, green, and blue channels;

[0024] Confirm the multi-channel image sensor mean value corresponding to each row in the target image reference area And the mean of the multi-channel image sensor corresponding to each column ;

[0025] By formula , and obtain the correction error coefficient corresponding to each row in the reference area of ​​the target image , and by the formula , and obtain the correction error coefficients corresponding to each column in the target image reference area .

[0026] Preferably, after determining the row and column correction error coefficients corresponding to the target image, the method further includes:

[0027] Confirm the row and column information corresponding to the image sensor in the target image, and calculate it through the formula , calculate the pixel correction error coefficient corresponding to the image sensor ,in, Expressed as weight coefficient.

[0028] Preferably, it also includes:

[0029] The pixel correction error coefficient corresponding to each image sensor is stored in the memory corresponding to each image sensor, and then each image sensor obtains a corrected image after calculating the pixel correction error coefficient in its own memory during the next shooting.

[0030] In a second aspect, the present application provides an image equalization calibration system, which adopts the following technical solution:

[0031] A calibration system for image equalization, comprising:

[0032] A target image confirmation module is used to confirm the target wide-field camera, receive image data collected by each image sensor corresponding to the target wide-field camera, combine each image data into a wide-field original image in a preset arrangement, and mark the wide-field original image as the target image;

[0033] A correction error coefficient analysis module, used for acquiring a reference area corresponding to the target image based on the target image, confirming a mean value of a multi-channel image sensor of the reference area corresponding to the target image, and confirming a row and column correction error coefficient corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image;

[0034] The correction module is used to correct the target image according to the row and column correction error coefficients corresponding to the target image, and then extract the corrected image.

[0035] In a third aspect, the present application provides a computer-readable storage medium storing instructions, which, when executed on a computer, enables the computer to execute any one of the above-described image equalization calibration methods.

[0036] In summary, this application includes the following beneficial technical effects:

[0037] The present invention provides an image equalization calibration method, which obtains a reference area corresponding to a target image, confirms a multi-channel image sensor mean value of the target image corresponding to the reference area, confirms a row and column correction error coefficient corresponding to the target image according to the multi-channel image sensor mean value of the target image corresponding to the reference area, corrects the target image according to the row and column correction error coefficient, and then extracts the corrected image, thereby effectively reducing the situation in which high-quality images cannot be collected due to uneven processing effects of wide-format photos composed of photos taken by multi-channel image sensors. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the method for calibrating image equalization according to an embodiment of the present application.

[0039] Figure 2 It is a system schematic diagram of image equalization calibration according to an embodiment of the present application. DETAILED DESCRIPTION

[0040] The following is combined with Figure 1-2 This application is described in further detail.

[0041] Example 1

[0042] The embodiment of the present application discloses a calibration method for image equalization.

[0043] Reference Figure 1 , a calibration method for image equalization, comprising the following steps:

[0044] S1. Confirm the target wide-field camera, receive the image data collected by each image sensor corresponding to the target wide-field camera, combine the image data into a wide-field original image in a preset arrangement, and mark the wide-field original image as the target image;

[0045] S2. Based on the target image, a reference area corresponding to the target image is obtained, and a mean value of a multi-channel image sensor corresponding to the reference area of ​​the target image is determined, and a row and column correction error coefficient corresponding to the target image is determined according to the mean value of the multi-channel image sensor corresponding to the reference area of ​​the target image;

[0046] S3. Correct the target image according to the row and column correction error coefficients corresponding to the target image, and then extract the corrected image.

[0047] By adopting the above technical scheme, by obtaining the reference area corresponding to the target image and confirming the multi-channel image sensor mean value of the reference area corresponding to the target image, and confirming the row and column correction error coefficients corresponding to the target image according to the multi-channel image sensor mean value of the reference area corresponding to the target image, the target image is corrected according to the row and column correction error coefficients corresponding to the target image, and then the corrected image is extracted, thereby effectively reducing the situation where high-quality images cannot be collected due to the uneven processing effects of wide-format photos composed of photos taken by multi-channel image sensors.

[0048] Specifically, receiving the image data collected by each image sensor corresponding to the target wide field camera also includes preprocessing the image data collected by each image sensor, including denoising and correcting distortion. By removing noise and correcting distortion, the image quality can be improved, making the image clearer and easier to analyze;

[0049] Preprocessing is an integral step in the image analysis process. Preprocessing can significantly affect the quality and reliability of the final results. Correct preprocessing steps can ensure optimal performance and efficiency.

[0050] Furthermore, the preset arrangement is set to an arrangement sequence number corresponding to each image sensor of the target wide-field camera.

[0051] It should be noted that obtaining the reference area corresponding to the target image based on the target image specifically includes:

[0052] Get the grayscale histogram corresponding to the target image ,in, Represented as grayscale, , Represented as the gray level in the target image is The number of pixels;

[0053] Analyze the grayscale histogram corresponding to the target image and confirm the preset brightness range Then calculate the uniformity evaluation coefficient corresponding to the preset brightness range , ,in, Indicates the preset brightness range The average number of pixels corresponding to

[0054] Adjust and traverse the preset brightness range, and then confirm the uniform evaluation coefficient corresponding to each traversal, and store the uniform evaluation coefficient corresponding to each traversal in an evaluation set, sort the uniform evaluation coefficients in the evaluation set according to a preset order, confirm the target uniform evaluation coefficient, and then obtain the target brightness range corresponding to the target uniform evaluation coefficient, confirm the corresponding target pixel in the target image based on the target brightness range, and confirm the distribution area corresponding to the target pixel, and then extract the distribution coefficient corresponding to each distribution area, and then confirm the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel.

[0055] Specifically, the preset brightness range The corresponding average number of pixels is calculated by the following formula:

[0056] .

[0057] It should be noted that confirming the distribution area corresponding to the target pixel, and then extracting the distribution coefficient corresponding to each distribution area, and then confirming the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel, specifically includes:

[0058] Obtain the distribution area corresponding to the target pixel and confirm the basic information corresponding to each distribution area, including the number of distribution areas and the number of pixels corresponding to each distribution area. ,in, It is represented by the number corresponding to each distribution area. , Indicated as number The number of pixels corresponding to the distribution area;

[0059] By formula , and the distribution coefficient corresponding to each distribution area is obtained ,in, Represents the number corresponding to each pixel, Expressed as In the distribution area The brightness value corresponding to each pixel is Expressed as The mean brightness of the pixels corresponding to the distribution area;

[0060] Get the distribution coefficient corresponding to each distribution area , the distribution coefficient corresponding to each distribution area With the preset distribution coefficient threshold Compare, if there is Less than , then the distribution area is stored in the reference area set;

[0061] The area corresponding to each distribution area in the reference area set is confirmed, the distribution area with the largest area in the reference area set is obtained, and then the distribution area is set as the reference area corresponding to the target image.

[0062] Specifically, by confirming the reference area corresponding to the target image, it can be ensured that the selected area is statistically representative of the entire image, thereby providing a stable reference point for color correction, exposure adjustment, and other image enhancement techniques. Secondly, the reference area can help reduce the impact of environmental factors, such as uneven lighting and shadows, so that the algorithm can more accurately extract and analyze image features. In addition, by reducing the variability of image data, the training efficiency and generalization ability of the machine learning model can be improved. Finally, confirming the reference area corresponding to the target image helps to improve the decision-making accuracy of the automated system.

[0063] Further, confirming the mean value of the multi-channel image sensor of the reference area corresponding to the target image, and confirming the row and column correction error coefficients corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image, specifically includes:

[0064] Confirm the position information corresponding to the reference area and use express;

[0065] By formula , confirm the multi-channel image sensor mean corresponding to the target image reference area ,in, They are respectively represented as the number of rows and columns corresponding to the pixels in the reference area, Respectively, they are represented as reference areas The positions correspond to the pixel values ​​of the red, green, and blue channels;

[0066] Confirm the multi-channel image sensor mean value corresponding to each row in the target image reference area And the mean of the multi-channel image sensor corresponding to each column ;

[0067] Specifically, the mean value of the multi-channel image sensor corresponding to each row in the target image reference area Calculated by the following formula:

[0068] ;

[0069] The mean value of the multi-channel image sensor corresponding to each column in the target image reference area Calculated by the following formula:

[0070] ;

[0071] By formula , and obtain the correction error coefficient corresponding to each row in the reference area of ​​the target image , and by the formula , and obtain the correction error coefficients corresponding to each column in the target image reference area .

[0072] Furthermore, after determining the row and column correction error coefficients corresponding to the target image, the method further includes:

[0073] Confirm the row and column information corresponding to the image sensor in the target image, and calculate it through the formula , calculate the pixel correction error coefficient corresponding to the image sensor ,in, Expressed as weight coefficient.

[0074] Furthermore, it also includes:

[0075] The pixel correction error coefficient corresponding to each image sensor is stored in the memory corresponding to each image sensor, and then each image sensor obtains a corrected image after calculating the pixel correction error coefficient in its own memory during the next shooting.

[0076] Example 2

[0077] The embodiment of the present application also discloses a calibration system for image equalization.

[0078] Reference Figure 2 , a calibration system for image equalization, comprising:

[0079] A target image confirmation module is used to confirm the target wide-field camera, receive image data collected by each image sensor corresponding to the target wide-field camera, combine each image data into a wide-field original image in a preset arrangement, and mark the wide-field original image as the target image;

[0080] A correction error coefficient analysis module, used for acquiring a reference area corresponding to the target image based on the target image, confirming a mean value of a multi-channel image sensor of the reference area corresponding to the target image, and confirming a row and column correction error coefficient corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image;

[0081] The correction module is used to correct the target image according to the row and column correction error coefficients corresponding to the target image, and then extract the corrected image.

[0082] The above contents are merely examples and explanations of the concept of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

[0083] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0084] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A calibration method for image equalization, characterized in that: The following steps are involved: Confirming a target wide-field-of-view camera, receiving image data collected by each image sensor corresponding to the target wide-field-of-view camera, combining each image data into a wide-field-of-view original image in a preset arrangement, and marking the wide-field-of-view original image as a target image; Acquire a reference area corresponding to the target image based on the target image, confirm a multi-channel image sensor mean value of the reference area corresponding to the target image, and confirm a row and column correction error coefficient corresponding to the target image according to the multi-channel image sensor mean value of the reference area corresponding to the target image; The target image is corrected according to the row and column correction error coefficients corresponding to the target image, and then the corrected image is extracted.

2. The image equalization calibration method according to claim 1, characterized in that: The preset arrangement is set to the arrangement sequence number corresponding to each image sensor of the target wide-field camera.

3. The image equalization calibration method according to claim 1, characterized in that: Acquiring a reference area corresponding to the target image based on the target image specifically includes: Get the grayscale histogram corresponding to the target image ,in, Represented as grayscale, , Represented as the gray level in the target image is The number of pixels; Analyze the grayscale histogram corresponding to the target image and confirm the preset brightness range Then calculate the uniformity evaluation coefficient corresponding to the preset brightness range , ,in, Indicates the preset brightness range The average number of pixels corresponding to Adjust and traverse the preset brightness range, and then confirm the uniform evaluation coefficient corresponding to each traversal, and store the uniform evaluation coefficient corresponding to each traversal in an evaluation set, sort the uniform evaluation coefficients in the evaluation set according to a preset order, confirm the target uniform evaluation coefficient, and then obtain the target brightness range corresponding to the target uniform evaluation coefficient, confirm the corresponding target pixel in the target image based on the target brightness range, and confirm the distribution area corresponding to the target pixel, and then extract the distribution coefficient corresponding to each distribution area, and then confirm the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel.

4. The image equalization calibration method according to claim 3, characterized in that: Confirming the distribution area corresponding to the target pixel, and then extracting the distribution coefficient corresponding to each distribution area, and then confirming the reference area corresponding to the target image based on the distribution coefficient corresponding to the target pixel, specifically includes: Obtain the distribution area corresponding to the target pixel and confirm the basic information corresponding to each distribution area, including the number of distribution areas and the number of pixels corresponding to each distribution area. ,in, It is represented by the number corresponding to each distribution area. , Indicated as number The number of pixels corresponding to the distribution area; By formula , and the distribution coefficient corresponding to each distribution area is obtained ,in, Represents the number corresponding to each pixel, Expressed as In the distribution area The brightness value corresponding to each pixel is Expressed as The mean brightness of the pixels corresponding to the distribution area; Get the distribution coefficient corresponding to each distribution area , the distribution coefficient corresponding to each distribution area With the preset distribution coefficient threshold Compare, if there is Less than , then the distribution area is stored in the reference area set; The area corresponding to each distribution area in the reference area set is confirmed, the distribution area with the largest area in the reference area set is obtained, and then the distribution area is set as the reference area corresponding to the target image.

5. The image equalization calibration method according to claim 4, characterized in that: Confirming the mean value of the multi-channel image sensor of the reference area corresponding to the target image, and confirming the row and column correction error coefficients corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image, specifically includes: Confirm the position information corresponding to the reference area and use express; By formula , confirm the multi-channel image sensor mean corresponding to the target image reference area ,in, They are respectively represented as the number of rows and columns corresponding to the pixels in the reference area, Respectively, they are represented as reference areas The positions correspond to the pixel values ​​of the red, green, and blue channels; Confirm the multi-channel image sensor mean value corresponding to each row in the target image reference area And the mean of the multi-channel image sensor corresponding to each column ; By formula , and obtain the correction error coefficient corresponding to each row in the reference area of ​​the target image , and by the formula , and obtain the correction error coefficients corresponding to each column in the target image reference area .

6. The image equalization calibration method according to claim 5, characterized in that: After confirming the row and column correction error coefficients corresponding to the target image, it also includes: Confirm the row and column information corresponding to the image sensor in the target image, and calculate it through the formula , calculate the pixel correction error coefficient corresponding to the image sensor ,in, Expressed as weight coefficient.

7. The image equalization calibration method according to claim 6, characterized in that: Also includes: The pixel correction error coefficient corresponding to each image sensor is stored in the memory corresponding to each image sensor, and then each image sensor obtains a corrected image after calculating the pixel correction error coefficient in its own memory during the next shooting.

8. A calibration system for image equalization, characterized in that: include: A target image confirmation module is used to confirm the target wide-field camera, receive image data collected by each image sensor corresponding to the target wide-field camera, combine each image data into a wide-field original image in a preset arrangement, and mark the wide-field original image as the target image; A correction error coefficient analysis module, used for acquiring a reference area corresponding to the target image based on the target image, confirming a mean value of a multi-channel image sensor of the reference area corresponding to the target image, and confirming a row and column correction error coefficient corresponding to the target image according to the mean value of the multi-channel image sensor of the reference area corresponding to the target image; The correction module is used to correct the target image according to the row and column correction error coefficients corresponding to the target image, and then extract the corrected image.

9. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute the image equalization calibration method as claimed in any one of claims 1 to 7.

Images