Image quality monitoring method and system of vehicle-mounted image acquisition sensor

Abstract

The invention relates to the technical field of image processing, and provides an image quality monitoring method and system for a vehicle-mounted image acquisition sensor, and the method comprises the steps: obtaining an original image in real time, and segmenting the original image into a plurality of sub-images; for each subimage, performing multi-scale feature analysis, local contrast analysis and information entropy analysis on the subimage to obtain a multi-scale feature index, a local contrast and an information entropy of the subimage, and generating a subimage quality index of the subimage based on the multi-scale feature index, the local contrast and the information entropy; based on a preset spatial weight distribution method, generating spatial weights of the sub-images; generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each sub-image and the spatial weight; and comparing the comprehensive quality index with a preset quality index, and triggering an image quality alarm mechanism when the comprehensive quality index is smaller than the preset quality index. The method is helpful for improving the comprehensiveness and accuracy of image quality evaluation.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to an image quality monitoring method and system for a vehicle-mounted image acquisition sensor. Background Art

[0002] With the rapid development of intelligent driving technology, on-board image acquisition sensors play a vital role in scenarios such as autonomous driving, driving records, and environmental perception. However, in practical applications, on-board image acquisition sensors face complex and changing environmental challenges, such as lighting changes, weather conditions, sensor hardware aging, and lens contamination. These factors may cause the quality of the acquired images to deteriorate, thus affecting subsequent image processing and analysis. Therefore, how to accurately monitor the image quality of on-board image acquisition sensors has become a key technical issue to ensure the safety and reliability of intelligent driving systems. At present, the image quality monitoring methods of on-board image acquisition sensors mainly rely on the evaluation of a single indicator (such as clarity or noise level), which is difficult to fully reflect the overall quality of the image. Summary of the invention

[0003] The present application provides an image quality monitoring method and system for a vehicle-mounted image acquisition sensor to solve the problems raised by the above-mentioned background technology.

[0004] In a first aspect, the present application provides an image quality monitoring method for a vehicle-mounted image acquisition sensor, comprising: Acquire an original image in real time, and divide the original image into multiple sub-images; For each of the sub-images, multi-scale feature analysis, local contrast analysis and information entropy analysis are performed on the sub-image to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and a sub-image quality index of the sub-image is generated based on the multi-scale feature index, the local contrast and the information entropy; Generating the spatial weight of each sub-image based on a preset spatial weight allocation method; Generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight; The comprehensive quality index is compared with a preset quality index, and when the comprehensive quality index is less than the preset quality index, an image quality alarm mechanism is triggered.

[0005] In a possible implementation manner, performing a multi-scale feature analysis on the sub-image to obtain a multi-scale feature index of the sub-image includes: Constructing a Gaussian pyramid corresponding to the sub-image; the resolution of each image in the Gaussian pyramid decreases from bottom to top, and the Gaussian pyramid includes a base number layer; Generate low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid; Generate a high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid; Calculating the edge strength of the middle layer image of the Gaussian pyramid by using the Sobel operator; The multi-scale feature index is generated based on the low-frequency energy, the high-frequency variance, and the edge strength.

[0006] In a possible implementation manner, the generating low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid includes: Obtaining the RGB color space of each pixel of the topmost image respectively; Generate low-frequency red energy, low-frequency blue energy and low-frequency green energy of the top-most image respectively based on the RGB color space of each pixel of the top-most image; The sum of the low-frequency red energy, the low-frequency blue energy and the low-frequency green energy is determined to be the low-frequency energy.

[0007] In a possible implementation manner, the generating the high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid includes: Obtaining the RGB color space of each pixel of the bottom layer image respectively; Generate a high-frequency red variance, a high-frequency blue variance, and a high-frequency green variance of the bottom layer image respectively based on the RGB color space of each pixel of the bottom layer image; The maximum value among the high-frequency red variance, the high-frequency blue variance and the high-frequency green variance is determined as the high-frequency variance.

[0008] In a possible implementation manner, calculating the edge strength of the intermediate layer image of the Gaussian pyramid by using a Sobel operator includes: Respectively obtain the RGB color space of each pixel of the intermediate layer image; Generate a red horizontal gradient matrix, a blue horizontal gradient matrix, a green horizontal gradient matrix, a red vertical gradient matrix, a blue vertical gradient matrix, and a green vertical gradient matrix of the intermediate layer image based on the RGB color space of each pixel of the intermediate layer image and the Sobel operator; Generate red edge intensity corresponding to each pixel of the intermediate layer image based on the red horizontal gradient matrix and the red vertical gradient matrix respectively; Generate blue edge intensity corresponding to each pixel of the intermediate layer image based on the blue horizontal gradient matrix and the blue vertical gradient matrix respectively; Generate green edge intensity corresponding to each pixel of the intermediate layer image based on the green horizontal gradient matrix and the green vertical gradient matrix respectively; For each pixel of the intermediate layer image, add the red edge intensity, the blue edge intensity and the green edge intensity corresponding to the pixel to obtain the edge intensity corresponding to the pixel; An average value of the edge strength corresponding to each pixel of the intermediate layer image is determined as the edge strength of the intermediate layer image.

[0009] In a possible implementation manner, performing local contrast analysis on the sub-image to obtain the local contrast includes: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate a red standard deviation, a blue standard deviation and a green standard deviation of the sub-image respectively based on the RGB color space of each pixel of the sub-image; The maximum value among the red standard deviation, the blue standard deviation and the green standard deviation is determined as the local contrast.

[0010] In a possible implementation manner, performing information entropy analysis on the sub-image to obtain information entropy includes: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate red information entropy, blue information entropy and green information entropy of the sub-image respectively based on the RGB color space of each pixel of the sub-image; Determine that the sum of the red information entropy, the blue information entropy and the green information entropy is the information entropy.

[0011] In a possible implementation, the generating of the spatial weights of the sub-images based on a preset spatial weight allocation method includes: Respectively obtaining the distance value between the center point of each sub-image and the center point of the original image, and calculating the sum of each distance value; For each of the sub-images, a ratio of a distance value corresponding to the sub-image to a sum of the distance values ​​is determined as a spatial weight corresponding to the sub-image.

[0012] In a second aspect, the present application provides an image quality monitoring system for a vehicle-mounted image acquisition sensor, comprising: An acquisition module, used for acquiring an original image in real time and dividing the original image into a plurality of sub-images; A first generating module is used for performing multi-scale feature analysis, local contrast analysis and information entropy analysis on each of the sub-images to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and generating a sub-image quality index of the sub-image based on the multi-scale feature index, the local contrast and the information entropy; A second generating module, used for respectively generating the spatial weight of each of the sub-images based on a preset spatial weight allocation method; A third generating module, used for generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight; The alarm module is used to compare the comprehensive quality index with a preset quality index and trigger an image quality alarm mechanism when the comprehensive quality index is less than the preset quality index.

[0013] The present application provides an image quality monitoring method and system for an on-vehicle image acquisition sensor, the method comprising: acquiring an original image in real time, and dividing the original image into a plurality of sub-images; for each of the sub-images, performing multi-scale feature analysis, local contrast analysis and information entropy analysis on the sub-image respectively, obtaining a multi-scale feature index, local contrast and information entropy of the sub-image, and generating a sub-image quality index of the sub-image based on the multi-scale feature index, the local contrast and the information entropy; generating a spatial weight of each of the sub-images based on a preset spatial weight allocation method; generating a comprehensive quality index of the original image based on the sub-image quality index and the spatial weight corresponding to each of the sub-images; comparing the comprehensive quality index with a preset quality index, and triggering an image quality alarm mechanism when the comprehensive quality index is less than the preset quality index. This method, firstly, can perform fine processing on the image by dividing the original image into multiple sub-images, thus overcoming the defect of the global evaluation method in the prior art that is insensitive to local quality changes. Then, multi-scale feature analysis, local contrast analysis and information entropy analysis are performed on each sub-image respectively to comprehensively capture the details, contrast and information volume characteristics of the image, thus significantly improving the comprehensiveness and accuracy of image quality evaluation. Finally, by comparing the comprehensive quality index with the preset quality index and triggering the alarm mechanism, image quality problems can be discovered in time and corresponding measures can be taken, thus providing a strong guarantee for the safety and reliability of the intelligent driving system, while improving the practicality and reliability of the on-board image acquisition sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0015] Figure 1 A schematic diagram of a flow chart of an image quality monitoring method for a vehicle-mounted image acquisition sensor provided in an embodiment of the present application; Figure 2 A schematic block diagram of the structure of an image quality monitoring system for a vehicle-mounted image acquisition sensor provided in an embodiment of the present application; Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0018] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0019] It should be further understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0020] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0021] See also Figure 1 , Figure 1A flow chart of an image quality monitoring method for a vehicle-mounted image acquisition sensor provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the image quality monitoring method of the vehicle-mounted image acquisition sensor provided in the embodiment of the present application includes steps S1 to S5.

[0022] Step S1, acquiring an original image in real time, and dividing the original image into multiple sub-images.

[0023] Step S2: for each of the sub-images, perform multi-scale feature analysis, local contrast analysis and information entropy analysis on the sub-image to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and generate a sub-image quality index of the sub-image based on the multi-scale feature index, the local contrast and the information entropy.

[0024] Step S3: Generate the spatial weight of each sub-image based on a preset spatial weight allocation method.

[0025] Step S4: generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight.

[0026] Step S5: compare the comprehensive quality index with a preset quality index, and trigger an image quality alarm mechanism when the comprehensive quality index is less than the preset quality index.

[0027] In this embodiment, it specifically includes: In the above step S1, during the driving process of the vehicle, the original image of the vehicle-mounted image acquisition sensor is acquired in real time, and the original image is divided into a plurality of sub-images.

[0028] In the above step S2, for each sub-image, first, a multi-scale feature analysis is performed on the sub-image to obtain a multi-scale feature index of the sub-image, and a local contrast analysis is performed on the sub-image to obtain a local contrast of the sub-image, and an information entropy analysis is performed on the sub-image to obtain an information entropy of the sub-image, and then, a weighted sum is performed on the multi-scale feature index, the local contrast and the information entropy to obtain a sub-image quality index of the sub-image.

[0029] In the above step S3, first, the distance values ​​between the center points of each sub-image and the center point of the original image are respectively obtained, and the sum of each distance value is calculated, and then, for each sub-image, the ratio of the distance value corresponding to the sub-image to the sum of the distance values ​​is determined as the spatial weight corresponding to the sub-image; wherein the center point of the original image is the center point of the circumscribed circle of the original image, and for each sub-image, the center point of the sub-image is the center point of the circumscribed circle of the sub-image.

[0030] In the above step S4, based on the spatial weights corresponding to the sub-images, the sub-image quality indexes corresponding to the sub-images are weighted and summed to obtain the comprehensive quality index of the original image.

[0031] In the above step S5, the comprehensive quality index is compared with a preset quality index, and when the comprehensive quality index is less than the preset quality index, an image quality alarm mechanism is triggered.

[0032] The method provided in this embodiment, firstly, by dividing the original image into multiple sub-images, it is possible to perform fine processing on the image, thereby overcoming the defect that the global evaluation method in the prior art is insensitive to local quality changes. Then, multi-scale feature analysis, local contrast analysis and information entropy analysis are performed on each sub-image respectively, so as to comprehensively capture the details, contrast and information volume characteristics of the image, and significantly improve the comprehensiveness and accuracy of image quality evaluation. Finally, by comparing the comprehensive quality index with the preset quality index and triggering the alarm mechanism, it is possible to timely discover image quality problems and take corresponding measures, thereby providing a strong guarantee for the safety and reliability of the intelligent driving system, and at the same time improving the practicality and reliability of the vehicle-mounted image acquisition sensor.

[0033] In some embodiments, performing multi-scale feature analysis on the sub-image to obtain a multi-scale feature index of the sub-image includes the following steps: Constructing a Gaussian pyramid corresponding to the sub-image; the resolution of each image in the Gaussian pyramid decreases from bottom to top, and the Gaussian pyramid includes a base number layer; Generate low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid; Generate a high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid; Calculating the edge strength of the middle layer image of the Gaussian pyramid by using the Sobel operator; The multi-scale feature index is generated based on the low-frequency energy, the high-frequency variance and the edge strength; specifically, the low-frequency energy, the high-frequency variance and the edge strength are weightedly summed to obtain the multi-scale feature index.

[0034] The method provided in this embodiment, firstly, by constructing a Gaussian pyramid of sub-images, it is possible to capture the multi-scale features of an image from different resolution levels, thereby comprehensively reflecting the overall structure and detail information of the image. Then, based on the top-level image of the Gaussian pyramid, low-frequency energy is generated to effectively characterize the overall brightness and structural distribution of the image, thereby providing a basic basis for image quality assessment. Secondly, based on the bottom-level image of the Gaussian pyramid, high-frequency variance is generated to accurately quantify the detail richness and noise level of the image, thereby enhancing the sensitivity of the assessment method to image details. Next, the edge strength of the intermediate layer image is calculated by the Sobel operator to further capture the clarity and edge information of the image, thereby improving the ability of the assessment method to judge the clarity of the image. Finally, by weighted summing the low-frequency energy, high-frequency variance and edge strength to generate a multi-scale feature index, a comprehensive assessment of the overall and local features of the image is achieved, thereby significantly improving the accuracy and comprehensiveness of image quality assessment, thereby providing reliable technical support for image quality monitoring of vehicle-mounted image acquisition sensors.

[0035] The step of generating the low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid comprises the following steps: Obtaining the RGB color space of each pixel of the topmost image respectively; Based on the RGB color space of each pixel of the topmost image, the low-frequency red energy, low-frequency blue energy and low-frequency green energy of the topmost image are respectively generated; wherein the low-frequency red energy is the average value of the sum of squares of the red values ​​of each pixel of the topmost image, the low-frequency blue energy is the average value of the sum of squares of the blue values ​​of each pixel of the topmost image, and the low-frequency green energy is the average value of the sum of squares of the green values ​​of each pixel of the topmost image, The sum of the low-frequency red energy, the low-frequency blue energy and the low-frequency green energy is determined to be the low-frequency energy.

[0036] It can be understood that the method for generating the low-frequency energy of the sub-image based on the top-level image of the Gaussian pyramid described above, firstly, by obtaining the RGB color space of each pixel of the top-level image of the Gaussian pyramid, can make full use of the color information of the image and provide multi-dimensional data support for the low-frequency energy calculation. Then, the low-frequency red energy, low-frequency blue energy and low-frequency green energy of the top-level image are calculated respectively, and the energy distribution characteristics of each color channel are effectively captured by calculating the square sum average, thereby enhancing the accuracy and stability of the low-frequency energy calculation. Secondly, the sum of the low-frequency red energy, the low-frequency blue energy and the low-frequency green energy is determined as the low-frequency energy, which comprehensively reflects the characteristics of the image in terms of overall brightness distribution and color structure, and provides comprehensive low-frequency information support for image quality assessment.

[0037] The step of generating the high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid comprises the following steps: Obtaining the RGB color space of each pixel of the bottom layer image respectively; Generate a high-frequency red variance, a high-frequency blue variance, and a high-frequency green variance of the bottom layer image respectively based on the RGB color space of each pixel of the bottom layer image; The maximum value among the high-frequency red variance, the high-frequency blue variance and the high-frequency green variance is determined as the high-frequency variance.

[0038] It can be understood that the method for generating the high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid described above, first, by obtaining the RGB color space of each pixel of the bottom-level image of the Gaussian pyramid, can make full use of the color information of the image and provide multi-dimensional data support for high-frequency variance calculation. Then, the high-frequency red variance, high-frequency blue variance and high-frequency green variance of the bottom-level image are calculated respectively, and the detail changes and noise levels of each color channel are effectively captured through variance calculation, thereby enhancing the accuracy and robustness of high-frequency variance calculation. Finally, the maximum value of the high-frequency red variance, the high-frequency blue variance and the high-frequency green variance is determined as the high-frequency variance, which can highlight the color channel with the richest details or the most significant noise in the image, thereby more accurately reflecting the detail performance and noise characteristics of the image.

[0039] The step of calculating the edge strength of the middle layer image of the Gaussian pyramid by using the Sobel operator comprises the following steps: Respectively obtain the RGB color space of each pixel of the intermediate layer image; Generate a red horizontal gradient matrix, a blue horizontal gradient matrix, a green horizontal gradient matrix, a red vertical gradient matrix, a blue vertical gradient matrix, and a green vertical gradient matrix of the intermediate layer image based on the RGB color space of each pixel of the intermediate layer image and the Sobel operator; Based on the red horizontal gradient matrix and the red vertical gradient matrix, respectively generate the red edge intensity corresponding to each pixel of the intermediate layer image; specifically, for each pixel, determine the red horizontal gradient value corresponding to the pixel in the red horizontal gradient matrix, determine the red vertical gradient value corresponding to the pixel in the red vertical gradient matrix, and determine the square root corresponding to the sum of the squares of the red horizontal gradient value and the red vertical gradient value as the red edge intensity of the pixel; Based on the blue horizontal gradient matrix and the blue vertical gradient matrix, respectively generate the blue edge intensity corresponding to each pixel of the intermediate layer image; specifically, for each pixel, determine the blue horizontal gradient value corresponding to the pixel in the blue horizontal gradient matrix, determine the blue vertical gradient value corresponding to the pixel in the blue vertical gradient matrix, and determine the square root corresponding to the sum of the squares of the blue horizontal gradient value and the blue vertical gradient value as the blue edge intensity of the pixel; Based on the green horizontal gradient matrix and the green vertical gradient matrix, the green edge intensity corresponding to each pixel of the intermediate layer image is respectively generated; specifically, for each pixel, the green horizontal gradient value corresponding to the pixel is determined in the green horizontal gradient matrix, the green vertical gradient value corresponding to the pixel is determined in the green vertical gradient matrix, and the square root corresponding to the sum of the squares of the green horizontal gradient value and the green vertical gradient value is determined as the green edge intensity of the pixel; For each pixel of the intermediate layer image, add the red edge intensity, the blue edge intensity and the green edge intensity corresponding to the pixel to obtain the edge intensity corresponding to the pixel; An average value of the edge strength corresponding to each pixel of the intermediate layer image is determined as the edge strength of the intermediate layer image.

[0040] It can be understood that the method for calculating the edge strength of the intermediate layer image of the Gaussian pyramid by using the Sobel operator described above, first, by obtaining the RGB color space of each pixel of the intermediate layer image of the Gaussian pyramid, the color information of the image can be fully utilized to provide multi-dimensional data support for edge strength calculation. Then, based on the Sobel operator, the horizontal gradient matrix and the vertical gradient matrix of the red, blue and green channels are respectively generated to effectively capture the edge change characteristics of each color channel in the horizontal and vertical directions, thereby enhancing the accuracy and comprehensiveness of edge detection. Secondly, by calculating the edge strength of each color channel, the edge information of the image in the red, blue and green channels can be accurately quantified. Then, the red, blue and green edge strengths corresponding to each pixel are added, which comprehensively reflects the overall edge characteristics of the image, further enhancing the comprehensiveness and accuracy of the edge strength calculation. Finally, by calculating the average value of the edge strength of each pixel, the edge strength of the intermediate layer image is obtained, thereby providing a reliable technical basis for image clarity evaluation.

[0041] In some embodiments, performing local contrast analysis on the sub-image to obtain the local contrast includes the following steps: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate a red standard deviation, a blue standard deviation and a green standard deviation of the sub-image respectively based on the RGB color space of each pixel of the sub-image; The maximum value among the red standard deviation, the blue standard deviation and the green standard deviation is determined as the local contrast.

[0042] The method provided in this embodiment, firstly, by acquiring the RGB color space of each pixel of the sub-image, can make full use of the color information of the image and provide multi-dimensional data support for the local contrast calculation; then, the red standard deviation, blue standard deviation and green standard deviation of the sub-image are calculated respectively, and the local brightness change characteristics of each color channel are effectively captured by the standard deviation calculation method, thereby enhancing the accuracy and robustness of the local contrast calculation; finally, the maximum value of the red standard deviation, the blue standard deviation and the green standard deviation is determined as the local contrast, which can highlight the color channel with the most significant contrast in the image, thereby more accurately reflecting the local detail performance of the image.

[0043] In some embodiments, performing information entropy analysis on the sub-image to obtain information entropy includes the following steps: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate red information entropy, blue information entropy and green information entropy of the sub-image respectively based on the RGB color space of each pixel of the sub-image; Determine that the sum of the red information entropy, the blue information entropy and the green information entropy is the information entropy.

[0044] The step of respectively generating the red information entropy of the sub-image based on the RGB color space of each pixel of the sub-image comprises: Determine the red value category based on the RGB color space of each pixel of the sub-image; for example, all red values ​​of each pixel of the sub-image include 125, 135, 128, and 215, then the red value category includes four categories: 125, 135, 128, and 215; For each of the red value categories, determining a probability that a ratio of the number of pixels corresponding to the red value category to the total number of pixels of the sub-image is the probability corresponding to the red value category; pass Obtain the red information entropy; wherein, is the red information entropy, For the The probability corresponding to the red value category, The number of red value categories is .

[0045] It should be noted that the method for generating the blue information entropy and the green information entropy can refer to the method for generating the red information entropy described above, which will not be repeated here.

[0046] The method provided in this embodiment, firstly, by acquiring the RGB color space of each pixel of the sub-image, can make full use of the color information of the image and provide multi-dimensional data support for information entropy calculation; then, the red information entropy, blue information entropy and green information entropy of the sub-image are calculated respectively, and the complexity and information quantity characteristics of each color channel are effectively captured by the information entropy calculation method, thereby enhancing the accuracy and comprehensiveness of the information entropy calculation; finally, the sum of the red information entropy, the blue information entropy and the green information entropy is determined as the information entropy of the sub-image, which can comprehensively reflect the characteristics of the image in terms of overall information quantity and complexity, thereby providing comprehensive information quantity support for image quality assessment.

[0047] See also Figure 2 , Figure 2 The structure schematic block diagram of the image quality monitoring system 100 of the vehicle-mounted image acquisition sensor provided in the embodiment of the present application is as follows: Figure 2 As shown, the image quality monitoring system 100 of the vehicle-mounted image acquisition sensor provided in the embodiment of the present application includes: The acquisition module 110 is used to acquire an original image in real time and divide the original image into multiple sub-images.

[0048] The first generating module 120 is used to perform multi-scale feature analysis, local contrast analysis and information entropy analysis on each of the sub-images to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and generate a sub-image quality index of the sub-image based on the multi-scale feature index, the local contrast and the information entropy.

[0049] The second generating module 130 is configured to generate the spatial weight of each sub-image based on a preset spatial weight allocation method.

[0050] The third generating module 140 is used to generate a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight.

[0051] The alarm module 150 is used to compare the comprehensive quality index with a preset quality index, and trigger an image quality alarm mechanism when the comprehensive quality index is less than the preset quality index.

[0052] It should be noted that, those skilled in the art can clearly understand that, for the sake of convenience and conciseness of description, the specific working process of the system and each module described above can refer to the process in the aforementioned image quality monitoring method embodiment of the vehicle-mounted image acquisition sensor, and will not be repeated here.

[0053] The image quality monitoring system 100 for the vehicle-mounted image acquisition sensor provided in the above embodiment can be implemented in the form of a computer program. The computer program can be used in Figure 3 The system is run on the terminal device 200 shown.

[0054] See also Figure 3 , Figure 3 The present invention provides a schematic block diagram of the structure of a terminal device 200 according to an embodiment of the present application. The terminal device 200 includes a processor 201 and a memory 202. The processor 201 and the memory 202 are connected via a device bus 203, wherein the memory 202 can include a non-volatile storage medium and an internal memory.

[0055] The non-volatile storage medium can store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 201, the processor 201 can execute any of the above-mentioned image quality monitoring methods for the vehicle-mounted image acquisition sensor.

[0056] The processor 201 is used to provide computing and control capabilities to support the operation of the entire terminal device 200.

[0057] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 201, the processor 201 can execute any of the above-mentioned image quality monitoring methods for the vehicle-mounted image acquisition sensor.

[0058] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the terminal device 200 involved in the scheme of the present application. The specific terminal device 200 can include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0059] It should be understood that the processor 201 can be a central processing unit (CPU), and the processor 201 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0060] In some embodiments, the processor 201 is used to run a computer program stored in the memory to implement the following steps: Acquire an original image in real time, and divide the original image into multiple sub-images; For each of the sub-images, multi-scale feature analysis, local contrast analysis and information entropy analysis are performed on the sub-image to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and a sub-image quality index of the sub-image is generated based on the multi-scale feature index, the local contrast and the information entropy; Generating the spatial weight of each sub-image based on a preset spatial weight allocation method; Generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight; The comprehensive quality index is compared with a preset quality index, and when the comprehensive quality index is less than the preset quality index, an image quality alarm mechanism is triggered.

[0061] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working process of the terminal device 200 described above can refer to the process of the image quality monitoring method of the aforementioned vehicle-mounted image acquisition sensor, and will not be repeated here.

[0062] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by one or more processors, the one or more processors implement the image quality monitoring method of the vehicle-mounted image acquisition sensor provided in the embodiment of the present application.

[0063] The computer-readable storage medium can be an internal storage unit of the terminal device 200 in the aforementioned embodiment, such as a hard disk or memory of the terminal device 200. The computer-readable storage medium can also be an external storage device of the terminal device 200, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped with the terminal device 200.

[0064] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for monitoring image quality of a vehicle-mounted image acquisition sensor, characterized in that: include: Acquire an original image in real time, and divide the original image into multiple sub-images; For each of the sub-images, multi-scale feature analysis, local contrast analysis and information entropy analysis are performed on the sub-image to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and a sub-image quality index of the sub-image is generated based on the multi-scale feature index, the local contrast and the information entropy; Generating the spatial weight of each sub-image based on a preset spatial weight allocation method; Generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight; The comprehensive quality index is compared with a preset quality index, and when the comprehensive quality index is less than the preset quality index, an image quality alarm mechanism is triggered.

2. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 1, characterized in that: Performing a multi-scale feature analysis on the sub-image to obtain a multi-scale feature index of the sub-image includes: Constructing a Gaussian pyramid corresponding to the sub-image; the resolution of each image in the Gaussian pyramid decreases from bottom to top, and the Gaussian pyramid includes a base number layer; Generate low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid; Generate a high-frequency variance of the sub-image based on the bottom-level image of the Gaussian pyramid; Calculating the edge strength of the middle layer image of the Gaussian pyramid by using the Sobel operator; The multi-scale feature index is generated based on the low-frequency energy, the high-frequency variance, and the edge strength.

3. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 2 is characterized in that: The step of generating the low-frequency energy of the sub-image based on the topmost image of the Gaussian pyramid comprises: Obtaining the RGB color space of each pixel of the topmost image respectively; Generate low-frequency red energy, low-frequency blue energy and low-frequency green energy of the top-most image respectively based on the RGB color space of each pixel of the top-most image; The sum of the low-frequency red energy, the low-frequency blue energy and the low-frequency green energy is determined to be the low-frequency energy.

4. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 2 is characterized in that: The step of generating the high frequency variance of the sub-image based on the bottom layer image of the Gaussian pyramid includes: Obtaining the RGB color space of each pixel of the bottom layer image respectively; Generate a high-frequency red variance, a high-frequency blue variance, and a high-frequency green variance of the bottom layer image respectively based on the RGB color space of each pixel of the bottom layer image; The maximum value among the high-frequency red variance, the high-frequency blue variance and the high-frequency green variance is determined as the high-frequency variance.

5. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 2, characterized in that: The step of calculating the edge strength of the intermediate layer image of the Gaussian pyramid by using the Sobel operator includes: Respectively obtain the RGB color space of each pixel of the intermediate layer image; Generate a red horizontal gradient matrix, a blue horizontal gradient matrix, a green horizontal gradient matrix, a red vertical gradient matrix, a blue vertical gradient matrix, and a green vertical gradient matrix of the intermediate layer image based on the RGB color space of each pixel of the intermediate layer image and the Sobel operator; Generate red edge intensity corresponding to each pixel of the intermediate layer image based on the red horizontal gradient matrix and the red vertical gradient matrix respectively; Generate blue edge intensity corresponding to each pixel of the intermediate layer image based on the blue horizontal gradient matrix and the blue vertical gradient matrix respectively; Generate green edge intensity corresponding to each pixel of the intermediate layer image based on the green horizontal gradient matrix and the green vertical gradient matrix respectively; For each pixel of the intermediate layer image, add the red edge intensity, the blue edge intensity and the green edge intensity corresponding to the pixel to obtain the edge intensity corresponding to the pixel; An average value of the edge strength corresponding to each pixel of the intermediate layer image is determined as the edge strength of the intermediate layer image.

6. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 1, characterized in that: Performing local contrast analysis on the sub-image to obtain local contrast includes: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate a red standard deviation, a blue standard deviation and a green standard deviation of the sub-image respectively based on the RGB color space of each pixel of the sub-image; The maximum value among the red standard deviation, the blue standard deviation and the green standard deviation is determined as the local contrast.

7. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 1, characterized in that: Performing information entropy analysis on the sub-image to obtain information entropy includes: Obtaining the RGB color space of each pixel of the sub-image respectively; Generate red information entropy, blue information entropy and green information entropy of the sub-image respectively based on the RGB color space of each pixel of the sub-image; Determine that the sum of the red information entropy, the blue information entropy and the green information entropy is the information entropy.

8. The image quality monitoring method of the vehicle-mounted image acquisition sensor according to claim 1, characterized in that: The generating of the spatial weights of the sub-images based on a preset spatial weight allocation method comprises: Respectively obtaining the distance value between the center point of each sub-image and the center point of the original image, and calculating the sum of each distance value; For each of the sub-images, a ratio of a distance value corresponding to the sub-image to a sum of the distance values ​​is determined as a spatial weight corresponding to the sub-image.

9. An image quality monitoring system for a vehicle-mounted image acquisition sensor, characterized in that: include: An acquisition module, used for acquiring an original image in real time and dividing the original image into a plurality of sub-images; A first generating module is used for performing multi-scale feature analysis, local contrast analysis and information entropy analysis on each of the sub-images to obtain a multi-scale feature index, local contrast and information entropy of the sub-image, and generating a sub-image quality index of the sub-image based on the multi-scale feature index, the local contrast and the information entropy; A second generating module, used for respectively generating the spatial weight of each of the sub-images based on a preset spatial weight allocation method; A third generating module, used for generating a comprehensive quality index of the original image based on the sub-image quality index corresponding to each of the sub-images and the spatial weight; The alarm module is used to compare the comprehensive quality index with a preset quality index and trigger an image quality alarm mechanism when the comprehensive quality index is less than the preset quality index.

Images