An AI camera image transmission optimization method for vehicles

Abstract

The present invention discloses an AI camera image transmission optimization method for a vehicle, and specifically relates to the field of image transmission optimization, including: performing AI recognition on an image being shot by a vehicle camera, and obtaining a magnification factor of the real-time transmission requirement of the shot image of the current driving scene; obtaining an original image shot by a camera within a set time and a transmitted image, comparing color richness feature deviations, and evaluating the degree of detail loss in image transmission; analyzing pixel motion features of a pixel-dense area of ​​an original image shot within a set time, and evaluating the degree of pixel motion chaos of the original image; analyzing driving speed features and road conditions, and evaluating the degree of driving control risk of the vehicle; comprehensively analyzing the degree of image transmission detail loss, pixel motion chaos, and vehicle driving control risk, evaluating the real-time transmission requirement of the current shot image, and optimizing and adjusting the transmission mode.

Description

Technical Field

[0001] The present invention relates to the technical field of image transmission optimization, and more specifically, to an AI camera image transmission optimization method for a vehicle. Background Art

[0002] A webcam generally has basic functions such as video recording / transmission and static image capture. After the lens captures the image, the photosensitive component circuit and control components in the camera process the image and convert it into a digital signal that can be recognized by the computer. It is then input into the computer through a parallel port or USB connection and the image is restored by the software.

[0003] Vehicle camera systems usually have high real-time requirements, quickly processing and displaying images to provide timely feedback and warnings to ensure driving safety. Limited by bandwidth and device processing performance, in some complex driving scenarios, the camera shooting accuracy does not match the system processing performance because the user uses a non-original standard camera, or the user manually sets the camera resolution too high, which reduces the image processing rate and leads to insufficient real-time performance of the captured image, resulting in limited functions of the assisted driving system including vehicle emergency braking, automatic cruise control and panoramic imaging, posing a potential safety hazard to vehicle driving safety.

[0004] How to provide a method to automatically adjust the shooting accuracy and transmission method to ensure the real-time transmission of image by evaluating the real-time requirements of the current vehicle's image transmission is an urgent problem to be solved.

[0005] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an AI camera image transmission optimization method for a vehicle to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for optimizing the transmission of an AI camera image for a vehicle comprises the following steps:

[0009] S1: Perform AI recognition on the images being captured by the vehicle camera, classify the driving scenes according to the recognition results, and obtain the real-time transmission requirement magnification factor of the captured images of the current driving scene;

[0010] S2: Obtain the original image and the transmitted image taken by the camera within a set time, compare the color richness feature deviation, and evaluate the degree of detail loss in image transmission;

[0011] S3: Acquire and analyze the pixel motion characteristics of the pixel-dense area of ​​the original image captured within a set time, and evaluate the degree of pixel motion chaos of the original image;

[0012] S4: Acquire driving information and road conditions within a set time interval, analyze driving speed characteristics and road conditions, and evaluate the driving control risk level of the vehicle;

[0013] S5: Comprehensively analyze the degree of detail loss in image transmission, the degree of pixel motion chaos, and the degree of risk in vehicle driving control, evaluate the real-time requirements for current image transmission, and optimize the transmission method.

[0014] In a preferred embodiment, the AI ​​recognition of the image being captured by the vehicle camera, the classification of the driving scene according to the recognition result, and the acquisition of the real-time transmission requirement magnification factor of the captured image of the current driving scene specifically include:

[0015] Obtain the roads and weather types allowed by vehicle specification standards, and set the real-time demand weights of camera transmission when driving on different types of roads and different types of weather according to actual conditions;

[0016] The roads that are allowed to pass according to the vehicle specification standards and the weather types that are allowed to pass are randomly arranged and combined, and each combination of road type and weather type is marked as a driving scene. The transmission real-time demand magnification coefficient array is established by taking the product of the transmission real-time demand weight of the road type and the weather type as the transmission real-time demand magnification coefficient of the captured image of the driving scene. , , is the transmission real-time demand magnification factor corresponding to the driving scenario of the combination of the i-th road category and the j-th weather type;

[0017] Perform AI recognition on the images taken by the vehicle camera in real time, classify the current driving scene of the vehicle according to the recognition results, and obtain the current driving scene's transmission real-time demand magnification coefficient array The corresponding transmission real-time demand amplification factor value.

[0018] In a preferred embodiment, the obtaining of the original image and the transmitted image captured by the camera within a set time, comparing the color richness feature deviation, and evaluating the degree of loss of image transmission details specifically includes:

[0019] Set the time interval T1, obtain the original image and the image information after transmission in the camera cache within the time T1, and analyze the color feature deviation between the original image and the image after transmission;

[0020] Get and read the original image, use the filter to smooth and denoise the original image, and select the RBG color space to create the RBG color histogram of the original image , is the value of the i-th color component in the color space of the original image, is the number of pixels corresponding to the i-th color component value in the color space of the original image;

[0021] Use the same method as above to build the RBG color histogram of the transferred image , and calculate the variance of the number of pixels corresponding to all color components in the color histogram of the original image and the transferred image at the same time;

[0022] By comparing and analyzing the color histogram deviation of the original image and the transmitted image, the color feature loss index is calculated by combining the variance of the number of pixels corresponding to each color component. The color feature loss index calculation formula is:

[0023]

[0024] In the formula, is the color feature loss index, is the number of pixels corresponding to the i-th color component of the color histogram of the j-th transmitted image, is the number of pixels corresponding to the i-th color component of the j-th original image color histogram, is the variance of the number of pixels corresponding to all color components of the color histogram of the j-th transmitted image, is the variance of the number of pixels corresponding to all color components of the color histogram of the jth original image, n is the number of color components of the color histogram of the original image and the image after transmission, and m is the number of comparison groups between the original image and the image after transmission.

[0025] In a preferred embodiment, the acquiring and analyzing the pixel motion characteristics of the pixel-dense area of ​​the original image captured within a set time and evaluating the degree of pixel motion chaos of the original image specifically includes:

[0026] Set the time interval T2, obtain all the original images taken within T2 according to the current shooting frame rate of the camera, extract the pixel features in each image, obtain the pixel-dense area, and compare the pixel motion of the pixel-dense area frame by frame, specifically:

[0027] Acquire the area in the image where the pixel density is greater than the set density, select the area with the largest pixel distribution area and mark it as the pixel-dense area, take the diagonal intersection of the minimum circumscribed rectangle of the irregular pixel distribution as the center point of the pixel-dense area, take the horizontal position of the center point of the pixel-dense area in the image as the abscissa, the vertical direction as the ordinate, and the shooting time sequence as the ordinal coordinate, and establish a three-dimensional rectangular coordinate system;

[0028] According to the orthogonal projection of the three-dimensional rectangular coordinate system on the x and y planes, a rectangular coordinate system for the motion plane of the center point of the time-sequential pixel-dense area is established. , the pixel motion chaos index of the image is calculated by comprehensively analyzing the movement distance of the central point of the time-order pixel-dense area, the slope of the line connecting the starting point and the final point relative to the initial movement trend, and the variance of the area change of the minimum circumscribed rectangle of the area. The calculation formula of the pixel motion chaos index is:

[0029]

[0030] In the formula, is the pixel motion chaos index, is the Euclidean distance from the coordinates of the center point of the i-th pixel-dense region to the coordinates of the center point of the i+1-th region, It is the slope of the line from the starting point of the center point of the pixel-dense area to the next point of the starting point. It is the slope of the line from the starting point to the final point at the center of the pixel-dense area. is the variance of the minimum bounding rectangle area change in the pixel-dense area, and n is the number of times the center point of the pixel-dense area moves.

[0031] In a preferred embodiment, the obtaining of driving information and road conditions within a set time interval, analyzing the driving speed characteristics and road conditions, and evaluating the driving control risk level of the vehicle specifically includes:

[0032] Set the time interval T3, obtain the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle during the T3 period, and identify the number of vehicles traveling within the set range captured by the AI ​​recognition camera;

[0033] Get the transmission real-time demand magnification factor of the current road, normalize the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle, and the number of vehicles traveling within the set range, and then perform weighted summation, calculate the product of the transmission real-time demand magnification factor of the current road, and mark it as the vehicle driving control risk index .

[0034] In a preferred embodiment, the comprehensive analysis of the degree of detail loss in image transmission, the degree of chaotic pixel motion, and the degree of risk in vehicle driving control, the evaluation of the real-time requirements for the current captured image transmission, and the optimization and adjustment of the transmission mode specifically include:

[0035] Obtain the color feature loss index and pixel motion chaos index of the image captured by the current vehicle camera, as well as the control risk index of the vehicle driving for comprehensive analysis, and calculate the transmission real-time demand index of the current captured image to evaluate the transmission real-time demand of the current captured image. The transmission real-time demand index calculation formula is:

[0036]

[0037] In the formula is the transmission real-time demand index, is the degree of pixel motion chaos, is the control risk index of vehicle driving, is the color feature loss index, , , They are , , The weight coefficient of

[0038] Set the vehicle camera transmission real-time demand index threshold. When the vehicle transmission real-time demand index is less than the transmission real-time demand index threshold, choose to increase the camera shooting resolution or reduce the compression level of the current vehicle camera image to ensure higher image quality.

[0039] When the vehicle transmission real-time demand index is greater than or equal to the transmission real-time demand index threshold, the transmission rate of the vehicle camera shooting is increased by reducing the camera shooting resolution or increasing the image compression level to ensure the real-time shooting.

[0040] Technical effects and advantages of the AI ​​camera image transmission optimization method for vehicles of the present invention:

[0041] 1. Perform AI recognition on the images being taken by the vehicle camera, classify the driving scenes according to the recognition results, set the initial amplification factor for the real-time transmission requirements of the camera image under different driving scenarios, and ensure that the transmission demand assessment is more targeted; compare the color richness feature deviation and evaluate the degree of detail loss in image transmission, which can reflect the real-time transmission adaptability of the current captured image; analyze the pixel motion characteristics of the pixel-dense areas of the original image captured within the set time, and evaluate the degree of pixel motion chaos of the original image, which can reflect the current vehicle speed and the complexity of the driving scene captured by the camera; analyze the driving speed characteristics and road conditions, and evaluate the degree of risk of vehicle driving control, which can reflect the driving safety level under the real-time transmission of the current camera-captured image.

[0042] 2. Through multi-angle monitoring of the degree of detail loss in image transmission, the degree of pixel motion chaos and the degree of control risk, it is evaluated whether the current real-time image transmission fails to meet the required standards. It emphasizes the attention to multiple factors of the real-time requirements of image transmission, and automatically adjusts the shooting resolution and transmission compression method in different scenarios of transmission real-time requirements to ensure the driving safety of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of an AI camera image transmission optimization method for a vehicle according to the present invention; DETAILED DESCRIPTION

[0044] 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 only 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.

[0045] Example 1

[0046] Figure 1 The present invention provides an AI camera image transmission optimization method for a vehicle, which comprises the following steps:

[0047] S1: Perform AI recognition on the images being captured by the vehicle camera, classify the driving scenes according to the recognition results, and obtain the real-time transmission requirement magnification factor of the captured images of the current driving scene;

[0048] S2: Obtain the original image and the transmitted image taken by the camera within a set time, compare the color richness feature deviation, and evaluate the degree of detail loss in image transmission;

[0049] S3: Acquire and analyze the pixel motion characteristics of the pixel-dense area of ​​the original image captured within a set time, and evaluate the degree of pixel motion chaos of the original image;

[0050] S4: Acquire driving information and road conditions within a set time interval, analyze driving speed characteristics and road conditions, and evaluate the driving control risk level of the vehicle;

[0051] S5: Comprehensively analyze the degree of detail loss in image transmission, the degree of pixel motion chaos, and the degree of risk in vehicle driving control, evaluate the real-time requirements for current image transmission, and optimize the transmission method.

[0052] Perform AI recognition on the images being taken by the vehicle camera, classify the driving scene according to the recognition results, and obtain the real-time transmission demand magnification factor of the captured image of the current driving scene. Specifically, the following are included:

[0053] Obtain the roads and weather types allowed by vehicle specification standards, and set the real-time demand weights of camera transmission when driving on different types of roads and different types of weather according to actual conditions;

[0054] The roads that are allowed to pass according to the vehicle specification standards and the weather types that are allowed to pass are randomly arranged and combined, and each combination of road type and weather type is marked as a driving scene. The transmission real-time demand magnification coefficient array is established by taking the product of the transmission real-time demand weight of the road type and the weather type as the transmission real-time demand magnification coefficient of the captured image of the driving scene. , , is the transmission real-time demand magnification factor corresponding to the driving scenario of the combination of the i-th road category and the j-th weather type;

[0055] Perform AI recognition on the images taken by the vehicle camera in real time, classify the current driving scene of the vehicle according to the recognition results, and obtain the current driving scene's transmission real-time demand magnification coefficient array The corresponding transmission real-time demand magnification coefficient values ​​specifically include:

[0056] Different road types and weather conditions have different requirements for the real-time transmission of images captured by vehicle cameras. For example, when driving on highways or crowded and narrow roads, the real-time transmission requirements for images will increase, while the real-time transmission requirements on open roads will decrease accordingly. When driving in bad weather, such as rainy days, the real-time transmission requirements for images will increase accordingly, while the real-time transmission requirements will decrease in normal weather. By combining the real-time transmission requirements for images of different road types and weather conditions, a more targeted transmission demand assessment method can be established.

[0057] Obtain the original image and the transmitted image taken by the camera within the set time, compare the color richness feature deviation, and evaluate the degree of detail loss in image transmission. Specifically include:

[0058] Set the time interval T1, obtain the original image and the image information after transmission in the camera cache within the time T1, and analyze the color feature deviation between the original image and the image after transmission;

[0059] Get and read the original image, use the filter to smooth and denoise the original image, and select the RBG color space to create the RBG color histogram of the original image , is the value of the i-th color component in the color space of the original image, is the number of pixels corresponding to the i-th color component value in the color space of the original image;

[0060] Use the same method as above to build the RBG color histogram of the transferred image , and calculate the variance of the number of pixels corresponding to all color components in the color histogram of the original image and the transferred image at the same time;

[0061] By comparing and analyzing the color histogram deviation of the original image and the transmitted image, the color feature loss index is calculated by combining the variance of the number of pixels corresponding to each color component. The color feature loss index calculation formula is:

[0062]

[0063] In the formula, is the color feature loss index, is the number of pixels corresponding to the i-th color component of the color histogram of the j-th transmitted image, is the number of pixels corresponding to the i-th color component of the j-th original image color histogram, is the variance of the number of pixels corresponding to all color components of the color histogram of the j-th transmitted image, is the variance of the number of pixels corresponding to all color components of the color histogram of the jth original image, n is the number of color components of the color histogram of the original image and the image after transmission, and m is the number of comparison groups between the original image and the image after transmission.

[0064] When using a lossy compression algorithm, the file size is reduced by discarding details and information, which will lead to a decrease in image quality. The higher the degree of compression, the more severe the detail loss. Color feature loss is a common type of compression detail loss. Therefore, by comparing the detail loss of the original image captured by the camera and the image after transmission, the current image real-time transmission adaptability can be obtained.

[0065] Obtaining and analyzing the pixel motion features of the pixel-dense area of ​​the original image taken within a set time, and evaluating the degree of pixel motion chaos of the original image specifically include:

[0066] Set the time interval T2, obtain all the original images taken within T2 according to the current shooting frame rate of the camera, extract the pixel features in each image, obtain the pixel-dense area, and compare the pixel motion of the pixel-dense area frame by frame, specifically:

[0067] Acquire the area in the image where the pixel density is greater than the set density, select the area with the largest pixel distribution area and mark it as the pixel-dense area, take the diagonal intersection of the minimum circumscribed rectangle of the irregular pixel distribution as the center point of the pixel-dense area, take the horizontal position of the center point of the pixel-dense area in the image as the abscissa, the vertical direction as the ordinate, and the shooting time sequence as the ordinal coordinate, and establish a three-dimensional rectangular coordinate system;

[0068] According to the orthogonal projection of the three-dimensional rectangular coordinate system on the x and y planes, a rectangular coordinate system for the motion plane of the center point of the time-sequential pixel-dense area is established. , the pixel motion chaos index of the image is calculated by comprehensively analyzing the movement distance of the central point of the time-order pixel-dense area, the slope of the line connecting the starting point and the final point relative to the initial movement trend, and the variance of the area change of the minimum circumscribed rectangle of the area. The calculation formula of the pixel motion chaos index is:

[0069]

[0070] In the formula, is the pixel motion chaos index, is the Euclidean distance from the coordinates of the center point of the i-th pixel-dense region to the coordinates of the center point of the i+1-th region, It is the slope of the line from the starting point of the center point of the pixel-dense area to the next point of the starting point. It is the slope of the line from the starting point to the final point at the center of the pixel-dense area. is the variance of the minimum bounding rectangle area change in the pixel-dense area, and n is the number of times the center point of the pixel-dense area moves.

[0071] Analyze the images taken during vehicle driving. Pixel-dense areas are generally road obstacles. When the pixel-dense area is larger than a certain area, it indicates that the obstacle may cause damage to the vehicle. By calculating the moving position of the center of the dense area and combining the image shooting frame rate, the current vehicle speed and the distance to the obstacle can be calculated. At the same time, according to the area change of the pixel-dense area and the slope deviation between the initial direction and the final direction, it can reflect whether the current vehicle is in an uneven road section, a turn or an irregular driving route. The degree of chaos of the pixel is comprehensively calculated as an evaluation indicator for the real-time image transmission requirements of the current driving scene.

[0072] Obtain driving information and road conditions within a set time interval, analyze driving speed characteristics and road conditions, and evaluate the driving control risk level of the vehicle, including:

[0073] Set the time interval T3, obtain the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle during the T3 period, and identify the number of vehicles traveling within the set range captured by the AI ​​recognition camera;

[0074] Get the transmission real-time demand magnification factor of the current road, normalize the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle, and the number of vehicles traveling within the set range, and then perform weighted summation, calculate the product of the transmission real-time demand magnification factor of the current road, and mark it as the vehicle driving control risk index .

[0075] The calculation formula of control risk index is:

[0076]

[0077] In the formula, is the control risk index, , , , They are the average frequency of acceleration change, the average amplitude of acceleration change and the maximum acceleration change, and the number of vehicles traveling within the set range. , , , They are , , , The weight coefficient of It is the transmission real-time demand amplification factor of the current driving scenario.

[0078] Frequent changes in acceleration during driving can reflect the vehicle's driving stability, while the maximum acceleration change can assess whether the vehicle has dangerous driving behaviors such as sudden speed increase or decrease or sudden braking. Combined with the number of driving vehicles within the monitoring range and the corresponding transmission real-time demand magnification factor of the road, a comprehensive assessment of the vehicle's current driving control risk level under the real-time transmission of image captured by the camera is used as the basis for assessing whether the current image transmission real-time meets the standard.

[0079] Comprehensively analyze the degree of detail loss in image transmission, the degree of pixel motion chaos, and the degree of risk in vehicle driving control, evaluate the real-time requirements for current captured image transmission, and optimize and adjust the transmission method, including:

[0080] Obtain the color feature loss index and pixel motion chaos index of the image captured by the current vehicle camera, as well as the control risk index of the vehicle driving for comprehensive analysis, and calculate the transmission real-time demand index of the current captured image to evaluate the transmission real-time demand of the current captured image. The transmission real-time demand index calculation formula is:

[0081]

[0082] In the formula is the transmission real-time demand index, is the degree of pixel motion chaos, is the control risk index of vehicle driving, is the color feature loss index, , , They are , , The weight coefficient of

[0083] Set the vehicle camera transmission real-time demand index threshold. When the vehicle transmission real-time demand index is less than the transmission real-time demand index threshold, choose to increase the camera shooting resolution or reduce the compression level of the current vehicle camera image to ensure higher image quality.

[0084] When the vehicle transmission real-time demand index is greater than or equal to the transmission real-time demand index threshold, the transmission rate of the vehicle camera shooting is increased by reducing the camera shooting resolution or increasing the image compression level to ensure the real-time shooting.

[0085] The transmission real-time demand index threshold is set by professional and technical personnel in this field based on the size of the vehicle camera transmission real-time demand index in the current driving scenario and other actual conditions such as vehicle specifications, types, and driving safety requirements, and will not be repeated here.

[0086] The camera shooting resolution and image compression level are automatically adjusted according to whether the current real-time transmission requirements are met, ensuring that users can choose to transmit higher quality images and improve the driving experience while meeting the real-time transmission requirements.

[0087] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0088] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0089] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0091] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0092] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0093] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0094] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0095] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0096] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing the image transmission of an AI camera for a vehicle, characterized in that: The steps include: S1: Perform AI recognition on the images being captured by the vehicle camera, classify the driving scenes according to the recognition results, and obtain the real-time transmission requirement magnification factor of the captured images of the current driving scene; S2: Obtain the original image and the transmitted image taken by the camera within a set time, compare the color richness feature deviation, and evaluate the degree of detail loss in image transmission; S3: Acquire and analyze the pixel motion characteristics of the pixel-dense area of ​​the original image captured within a set time, and evaluate the degree of pixel motion chaos of the original image; S4: Acquire driving information and road conditions within a set time interval, analyze driving speed characteristics and road conditions, and evaluate the driving control risk level of the vehicle; S5: Comprehensively analyze the degree of detail loss in image transmission, the degree of pixel motion chaos, and the degree of vehicle driving control risk, evaluate the real-time requirements for current image transmission, and optimize the transmission method; The AI ​​recognition of the image being captured by the vehicle camera, classification of the driving scene according to the recognition result, and acquisition of the real-time transmission demand magnification factor of the captured image of the current driving scene specifically include: Obtain the roads and weather types allowed by vehicle specification standards, and set the real-time demand weights of camera transmission when driving on different types of roads and different types of weather according to actual conditions; The roads that are allowed to pass according to the vehicle specification standards and the weather types that are allowed to pass are randomly arranged and combined, and each combination of road type and weather type is marked as a driving scene. The transmission real-time demand magnification coefficient array is established by taking the product of the transmission real-time demand weight of the road type and the weather type as the transmission real-time demand magnification coefficient of the captured image of the driving scene. , , is the transmission real-time demand magnification factor corresponding to the driving scenario of the combination of the i-th road category and the j-th weather type; Perform AI recognition on the images taken by the vehicle camera in real time, classify the current driving scene of the vehicle according to the recognition results, and obtain the current driving scene's transmission real-time demand magnification coefficient array The corresponding transmission real-time demand magnification factor value; The obtaining of the original image and the transmitted image captured by the camera within a set time, comparing the color richness feature deviation, and evaluating the degree of loss of image transmission details specifically includes: Set the time interval T1, obtain the original image and the image information after transmission in the camera cache within the time T1, and analyze the color feature deviation between the original image and the image after transmission; Get and read the original image, use the filter to smooth and denoise the original image, and select the RBG color space to create the RBG color histogram of the original image , is the value of the i-th color component in the color space of the original image, is the number of pixels corresponding to the i-th color component value in the color space of the original image; Use the same method as above to build the RBG color histogram of the transferred image , and calculate the variance of the number of pixels corresponding to all color components in the color histogram of the original image and the transferred image at the same time; By comparing and analyzing the color histogram deviation of the original image and the transmitted image, the color feature loss index is calculated by combining the variance of the number of pixels corresponding to each color component. The color feature loss index calculation formula is: ； In the formula, is the color feature loss index, is the number of pixels corresponding to the i-th color component of the color histogram of the j-th transmitted image, is the number of pixels corresponding to the i-th color component of the j-th original image color histogram, is the variance of the number of pixels corresponding to all color components of the color histogram of the j-th transmitted image, is the variance of the number of pixels corresponding to all color components of the color histogram of the jth original image, n is the number of color components of the color histogram of the original image and the image after transmission, and m is the number of comparison groups between the original image and the image after transmission; The obtaining and analyzing the pixel motion characteristics of the pixel-dense area of ​​the original image captured within a set time and evaluating the degree of pixel motion chaos of the original image specifically includes: Set the time interval T2, obtain all the original images taken within T2 according to the current shooting frame rate of the camera, extract the pixel features in each image, obtain the pixel-dense area, and compare the pixel motion of the pixel-dense area frame by frame, specifically: Acquire the area in the image where the pixel density is greater than the set density, select the area with the largest pixel distribution area and mark it as the pixel-dense area, take the diagonal intersection of the minimum circumscribed rectangle of the irregular pixel distribution as the center point of the pixel-dense area, take the horizontal position of the center point of the pixel-dense area in the image as the abscissa, the vertical direction as the ordinate, and the shooting time sequence as the ordinal coordinate, and establish a three-dimensional rectangular coordinate system; According to the orthogonal projection of the three-dimensional rectangular coordinate system on the x and y planes, a rectangular coordinate system for the motion plane of the center point of the time-sequential pixel-dense area is established. , the movement distance of the central point of the time-order pixel-dense area, the slope of the line connecting the starting point and the final point relative to the initial movement trend, and the variance of the minimum circumscribed rectangular area change of the pixel-dense area are comprehensively analyzed to calculate the pixel motion chaos index of the image. The calculation formula of the pixel motion chaos index is: ； In the formula, is the pixel motion chaos index, is the Euclidean distance from the coordinates of the center point of the i-th pixel-dense region to the coordinates of the center point of the i+1-th region, It is the slope of the line from the starting point of the center point of the pixel-dense area to the next point of the starting point. It is the slope of the line from the starting point to the final point at the center of the pixel-dense area. is the variance of the area change of the minimum bounding rectangle of the pixel-dense area, and n is the number of times the center point of the pixel-dense area moves; The obtaining of driving information and road conditions within a set time interval, analyzing the driving speed characteristics and road conditions, and evaluating the driving control risk level of the vehicle specifically includes: Set the time interval T3, obtain the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle during the T3 period, and identify the number of vehicles traveling within the set range captured by the AI ​​recognition camera; Get the transmission real-time demand magnification factor of the current road, normalize the average acceleration change frequency, average acceleration change amplitude and maximum acceleration change of the vehicle, and the number of vehicles traveling within the set range, and then perform weighted summation, calculate the product with the transmission real-time demand magnification factor of the current road, and mark it as the vehicle driving control risk index ; The comprehensive analysis of the degree of detail loss in image transmission, the degree of chaotic pixel motion, and the degree of risk in vehicle driving control, assessing the real-time requirements for the current captured image transmission, and optimizing and adjusting the transmission method specifically include: Obtain the color feature loss index and pixel motion chaos index of the image captured by the current vehicle camera, as well as the control risk index of the vehicle driving for comprehensive analysis, and calculate the transmission real-time demand index of the current captured image to evaluate the transmission real-time demand of the current captured image. The transmission real-time demand index calculation formula is: ； In the formula is the transmission real-time demand index, is the degree of pixel motion chaos, is the control risk index of vehicle driving, is the color feature loss index, , , They are , , The weight coefficient of Set the vehicle camera transmission real-time demand index threshold. When the vehicle transmission real-time demand index is less than the transmission real-time demand index threshold, choose to increase the camera shooting resolution or reduce the compression level of the current vehicle camera image to ensure higher image quality. When the vehicle transmission real-time demand index is greater than or equal to the transmission real-time demand index threshold, the transmission rate of the vehicle camera shooting is increased by reducing the camera shooting resolution or increasing the image compression level to ensure the real-time shooting.

Images