Face Image Quality Assessment Method, System and Computer Readable Storage Medium

Abstract

The present invention discloses a method, a system and a computer-readable storage medium for face image quality assessment, which relates to the technical field of face image quality assessment. Specifically, it includes the following steps: when it is determined that there is an artifact effect, the face image captured by the camera in real time is evenly divided into several regions by a facial key point detection method; the artifact quality assessment information of each region of the face image is obtained in real time, the quality influence degree of the artifact on each region is evaluated, and each region is divided into a high-influence region, a medium-influence region and a low-influence region according to the evaluation result; according to the division results of each region of the face image, a dynamic intervention mechanism is constructed, and corresponding intervention measures are taken for the high-influence region, the medium-influence region and the low-influence region respectively. The present invention solves the problem of misjudgment in face image quality assessment caused by artifact interference in a dynamic light source environment, and realizes the accurate assessment and efficient dynamic optimization of image quality.

Description

Technical Field

[0001] The present invention relates to the technical field of face image quality assessment, and specifically to a face image quality assessment method, system and computer-readable storage medium. Background Art

[0002] A face image refers to a digital image containing human facial features obtained by an image acquisition device (such as a camera, webcam, mobile phone, etc.). These images are usually used in application fields such as face recognition, emotion analysis, behavior monitoring, security monitoring, etc. Due to the influence of various factors such as illumination, shooting angle, resolution, occlusion, and noise, face images may have problems such as blurring, distortion, and low contrast, which seriously affect the accuracy of subsequent face recognition and analysis. Therefore, it is particularly important to evaluate the quality of face images. Quality assessment can help us quantify the clarity, contrast, noise level, and integrity of face features of the image, providing a more reliable and accurate input for subsequent tasks such as face recognition and emotion analysis. By evaluating the quality of the image, we can timely detect and correct defects in the image, thereby improving the stability and accuracy of the recognition system. In addition, quality assessment can also provide feedback for automated systems, optimize the image acquisition process, and ensure that the final input data obtained is of high quality. Therefore, image quality assessment is not only a key step in improving the effect of face image processing tasks, but also the basis for many computer vision systems to operate precisely and efficiently.

[0003] Existing face image quality assessment technologies usually combine traditional image processing methods and deep learning technologies, and use various methods to evaluate image quality. Traditional methods judge image quality by analyzing low-level features of the image, such as clarity, contrast, noise, brightness, etc. Specifically, clarity assessment usually uses edge detection algorithms (such as Sobel operator or Laplacian operator) to analyze details in the image, noise assessment identifies the noise level by calculating the standard deviation or mean of the image, and contrast assessment determines the contrast of the image by calculating the brightness difference in the local area. At the same time, deep learning methods automatically learn high-level and low-level features in the image by training a convolutional neural network (CNN) and output a quality score, which can comprehensively consider all aspects of the image and perform end-to-end quality prediction. In addition, generative adversarial networks (GANs) are also used to evaluate image quality, by comparing the differences between the generated high-quality images and the original images to identify quality problems in the images. Based on these technologies, the evaluation process usually includes feature extraction, quality scoring and analysis of the image, so as to provide reliable quality guarantee for subsequent face recognition or analysis tasks. Generally speaking, these technologies can accurately quantify the quality problems of the image and provide effective feedback for automated processing.

[0004] The existing technologies have the following deficiencies:

[0005] In the dynamic advertisement display scenario within a large shopping mall, a camera captures real-time facial images for user identification. Dynamic light sources (such as neon lights or reflected light) can generate high-reflection light in certain facial areas (such as the forehead) and propagate to adjacent areas (such as around the eyes), forming an artifact effect. This artifact is manifested as unnatural brightness transitions or patchy texture interferences. This is because dynamic light reflection not only changes the light intensity in the highlighted area but also affects the texture features and brightness distribution in adjacent areas through light propagation, causing the true features in these areas to be covered or distorted. Since the intensity and position of the artifact change dynamically with the light source angle and brightness, existing facial image quality assessment techniques rely on independent brightness, contrast, or texture analysis of each region and cannot accurately distinguish the difference between the artifact effect and defects in the true image features, resulting in confused quality assessment results. In this case, the system may misjudge the artifact area as a quality anomaly and take unnecessary intervention measures (such as adjusting exposure or increasing fill light), which instead interferes with the overall image quality. At the same time, the areas covered by the artifact effect (such as the eye texture) may be misjudged as abnormal areas, reducing the accuracy of key feature extraction by the face recognition algorithm and ultimately affecting the recognition efficiency and the accuracy of advertising placement decisions.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide a facial image quality assessment method, system, and computer-readable storage medium to solve the problems in the above background art.

[0008] To achieve the above object, the present invention provides the following technical solution: A facial image quality assessment method, specifically including the following steps:

[0009] In a dynamic light source environment, perform artifact effect detection on the facial images captured in real time by the camera. By analyzing the brightness distribution uniformity, abnormal gradient direction, and texture feature changes of the image, determine whether there is an artifact effect;

[0010] In the case where it is determined that there is an artifact effect, evenly divide the facial images captured in real time by the camera into several regions through a facial key point detection method;

[0011] Obtain the artifact quality assessment information of each region of the facial image in real time, and analyze it after obtaining. Evaluate the degree of quality impact of the artifact on each region, and divide each region into a high-impact region, a medium-impact region, and a low-impact region according to the evaluation results;

[0012] Construct a dynamic intervention mechanism based on the division results of each region of the face image, and perform corresponding intervention measures on the high-impact region, medium-impact region, and low-impact region respectively;

[0013] During the process of the dynamic intervention mechanism intervening in each region, obtain the intervention feedback information of each region in real time, and analyze it after obtaining, evaluate whether the intervention effect of the dynamic intervention mechanism in each region can meet the expectations, and optimize the dynamic intervention mechanism according to the evaluation results;

[0014] Perform real-time monitoring and comprehensive analysis on the artifact quality evaluation information, intervention feedback information, intervention records, and image quality restoration effects obtained during the artifact adjustment and evaluation process, continuously improve the artifact suppression method and dynamic intervention strategy, and optimize the face image quality evaluation process.

[0015] Preferably, obtain the artifact quality evaluation information of each region of the face image in real time, and analyze it after obtaining, evaluate the degree of quality impact of the artifact on each region, and divide each region into a high-impact region, a medium-impact region, and a low-impact region according to the evaluation results. What are the specific steps:

[0016] Obtain the artifact quality evaluation information of each region of the face image in real time, and perform preprocessing after obtaining;

[0017] Extract the brightness distribution characteristic information and texture structure characteristic information in the artifact quality evaluation information of each region of the preprocessed face image, and analyze them after extraction, and generate the brightness balance deviation index and detail texture damage index of each region respectively;

[0018] Construct an artifact impact evaluation model for the generated brightness balance deviation index and detail texture damage index of each region, generate the artifact impact coefficient of each region, and analyze it after generating, evaluate the degree of quality impact of the artifact on each region, and divide each region into a high-impact region, a medium-impact region, and a low-impact region according to the evaluation results.

[0019] Preferably, the acquisition logic of the brightness balance deviation index and detail texture damage index of each region is as follows:

[0020] Extract the brightness distribution characteristic information in the artifact quality evaluation information of each region of the preprocessed face image, specifically including the average value, variance of the brightness values of all pixel points in each region of the face image at different times within a period of time, and the absolute average difference between the brightness values of the edge pixels in each region and the adjacent edge pixels, and calibrate them respectively as 、 and , Indicates within a period of time At the moment, the face image is the The average brightness value of all pixel points in a region represents within a period of time the variance of the brightness values of all pixel points in the region of the face image at the moment; represents the absolute average difference between the brightness values of the edge pixels and the neighborhood edge pixels in the , , and are all positive integers;

[0021] Calculate the brightness balance deviation index of each region. The specific calculation formula is as follows:

[0022]

[0023] In the formula, is the brightness balance deviation index of the region;

[0024] Extract the texture structure characteristic information in the artifact quality evaluation information of each region of the pre - processed face image, specifically including the average change rate, amplitude, and direction distribution variance of the pixel gradients in each region of the face image at different moments within a period of time, and label them as , and , represents the average change rate of the pixel gradients in the region of the face image at the moment within a period of time, represents the amplitude of the pixel gradients in the region of the face image at the moment within a period of time, represents the direction distribution variance of the pixel gradients in the region of the face image at the moment within a period of time;

[0025] Calculate the detailed texture damage index of each region. The specific calculation formula is as follows:

[0026]

[0027] In the formula, is the detailed texture damage index of the region.

[0028] Preferably, for the generated brightness balance deviation index of each region and the detailed texture damage index Construct an artifact impact evaluation model, and generate the artifact impact coefficients of each region through weighted summation and compare the generated artifact impact coefficients of each region with the pre-set threshold range of the artifact impact coefficient to evaluate the quality impact degree of the artifact on each region according to the comparison result, and divide each region into a high-impact region, a medium-impact region, and a low-impact region according to the evaluation result. The specific comparison analysis and division are as follows:

[0029] If , the quality impact degree of the artifact on this region is low, then this region is divided into a low-impact region;

[0030] If , the quality impact degree of the artifact on this region is medium, then this region is divided into a medium-impact region;

[0031] If , the quality impact degree of the artifact on this region is high, then this region is divided into a high-impact region.

[0032] Preferably, according to the division results of each region of the face image, construct a dynamic intervention mechanism, specifically: according to the division results of the high-impact region, the medium-impact region, and the low-impact region, set different intervention measure parameters respectively to form a dynamic intervention mechanism; this dynamic intervention mechanism is based on the artifact impact coefficient of each region, and through pre-set rules, automatically determines the adjustment methods and amplitudes of brightness adjustment, texture enhancement, and edge smoothing;

[0033] Perform corresponding intervention measures on the high-impact region, the medium-impact region, and the low-impact region respectively, specifically: in the high-impact region, use the artifact suppression enhancement parameters in the dynamic intervention mechanism to significantly adjust the brightness distribution, strengthen texture enhancement, and optimize edge smoothing; in the medium-impact region, use the artifact medium adjustment parameters in the dynamic intervention mechanism to moderately optimize the brightness distribution and texture characteristics; in the low-impact region, maintain the artifact stable parameters in the dynamic intervention mechanism without adjustment, and only monitor the region status to maintain quality stability.

[0034] Preferably, during the process of the dynamic intervention mechanism intervening in each region, obtain the intervention feedback information of each region in real time, and analyze it after obtaining, evaluate whether the intervention effect of this dynamic intervention mechanism in each region can meet the expectation, and optimize the dynamic intervention mechanism according to the evaluation result, specifically including the following steps:

[0035] During the process of the dynamic intervention mechanism intervening in each region, obtain the intervention feedback information of each region in real time, and perform preprocessing after obtaining;

[0036] Extract the brightness adjustment feedback information and texture optimization feedback information from the intervention feedback information of each preprocessed region, and analyze them after acquisition to generate the brightness adjustment deviation index and texture recovery consistency index for each region respectively;

[0037] Construct an intervention effect evaluation model for the generated brightness adjustment deviation index and texture recovery consistency index of each region, generate the intervention coefficient of each region, compare the generated intervention coefficient of each region with the pre-set intervention coefficient threshold of each region, evaluate whether the intervention effect of this dynamic intervention mechanism in each region can meet the expectation according to the comparison result, and optimize the dynamic intervention mechanism according to the evaluation result.

[0038] Preferably, the acquisition logic of the brightness adjustment deviation index and texture recovery consistency index of each region is as follows:

[0039] Extract the brightness adjustment feedback information from the intervention feedback information of each preprocessed region, specifically including the average value, variance of the pixel brightness values of each region at different times within a period of time after the intervention, and the pre-set average value of the pixel brightness values of each region after the intervention, and calibrate them respectively as 、 and , represents the average value of the pixel brightness values of the th time within a period of time after the intervention in the th region, represents the variance of the pixel brightness values of the th time within a period of time after the intervention in the th region, represents the pre-set average value of the pixel brightness values of the th region after the intervention, , , and are all positive integers;

[0040] Calculate the brightness adjustment deviation index of each region. The specific calculation formula is as follows:

[0041]

[0042] In the formula, is the brightness adjustment deviation index of the th region;

[0043] Extract the texture optimization feedback information from the intervention feedback information of each preprocessed region, specifically including the average amplitude, direction distribution variance of the pixel gradients of each region at different times within a period of time after the intervention, and the pre-set average amplitude of the pixel gradients of each region after the intervention, and calibrate them respectively as , and , represents the average magnitude of the pixel gradient in the th moment within a period of time after the intervention for the th region, represents the variance of the direction distribution of the pixel gradient in the th moment within a period of time after the intervention for the th region, represents the preset average magnitude of the pixel gradient in the th region after the intervention;

[0044] Calculate the texture recovery consistency index for each region. The specific calculation formula is as follows:

[0045]

[0046] In the formula, is the texture recovery consistency index for the th region.

[0047] Preferably, for the brightness adjustment deviation index and the texture recovery consistency index generated for each region, construct an intervention effect evaluation model, generate the intervention coefficient for each region through weighted summation, and compare the generated intervention coefficient for each region with the preset intervention coefficient threshold for each region. According to the comparison result, evaluate whether the intervention effect of the dynamic intervention mechanism in each region can meet the expectation, and optimize the dynamic intervention mechanism according to the evaluation result. The specific comparison and analysis are as follows:

[0048] If , the intervention effect of the dynamic intervention mechanism in this region can meet the expectation, and there is no need to optimize the dynamic intervention mechanism;

[0049] If , the intervention effect of the dynamic intervention mechanism in this region cannot meet the expectation, and it is necessary to optimize the dynamic intervention mechanism, specifically including: increasing the brightness adjustment intensity of this region to reduce the brightness distribution deviation; increasing the parameter weight of texture enhancement to strengthen the contrast of details and the high-frequency texture characteristics; combining the real-time intervention feedback data of this region to dynamically adjust the intervention parameters and rules, and optimizing the execution process of the intervention mechanism.

[0050] Preferably, a face image quality assessment system includes an artifact effect detection module, a region division and calibration module, an artifact quality assessment module, a dynamic intervention execution module, an intervention effect optimization module, and a comprehensive monitoring and improvement module;

[0051] An artifact effect detection module that, in a dynamic light source environment, detects artifact effects on face images captured in real time by a camera. By analyzing the brightness distribution uniformity, abnormal gradient directions, and texture feature changes in the images, it determines whether there are artifact effects;

[0052] A region division and calibration module that, when it is determined that there are artifact effects, evenly divides the face images captured in real time by the camera into several regions through facial key point detection methods;

[0053] An artifact quality assessment module that obtains real-time artifact quality assessment information for each region of the face image, analyzes it after obtaining, evaluates the degree of quality impact of the artifacts on each region, and divides each region into high-impact regions, medium-impact regions, and low-impact regions according to the evaluation results;

[0054] A dynamic intervention execution module that, according to the division results of each region of the face image, constructs a dynamic intervention mechanism and performs corresponding intervention measures on the high-impact regions, medium-impact regions, and low-impact regions respectively;

[0055] An intervention effect optimization module that, during the process of the dynamic intervention mechanism intervening in each region, obtains real-time intervention feedback information for each region, analyzes it after obtaining, evaluates whether the intervention effect of the dynamic intervention mechanism in each region can meet the expectations, and optimizes the dynamic intervention mechanism according to the evaluation results;

[0056] A comprehensive monitoring and improvement module that conducts real-time monitoring and comprehensive analysis on the artifact quality assessment information, intervention feedback information, intervention records, and image quality restoration effects obtained during the artifact adjustment and assessment process, continuously improves the artifact suppression method and dynamic intervention strategy, and optimizes the face image quality assessment process.

[0057] Preferably, a computer-readable storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the face image quality assessment method.

[0058] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0059] 1. Through the artifact effect detection module of the present invention, the system can judge the existence of artifacts in real time, accurately identify the regions where artifacts are generated and their influence ranges, and avoid misjudgment of image features caused by artifacts in the prior art. The region division and artifact quality assessment module uses the brightness balance deviation index and the detail texture damage index to evaluate the specific impact of artifacts on each region in a quantitative manner, which not only improves the scientificity and accuracy of the assessment, but also provides a clear reference basis for the subsequent dynamic intervention mechanism, solving the problem of confused region quality assessment in the prior art.

[0060] 2. By implementing targeted intervention measures for high - impact areas, medium - impact areas, and low - impact areas respectively, the present invention can effectively improve the image quality of the artifact area. The high - impact area significantly improves the quality through enhanced brightness adjustment, texture enhancement, and edge smoothing, while the medium - impact and low - impact areas adopt moderate adjustment or monitoring to avoid waste of resources. This hierarchical intervention method significantly improves the processing efficiency and ensures fine - grained control of image quality, solving the problem of overall quality interference caused by unnecessary global adjustment in the prior art. The feedback mechanism during the intervention process further enhances the self - adaptability and accuracy of dynamic intervention through real - time monitoring and optimization, enabling the system to maintain excellent performance even under complex light source conditions.

[0061] 3. The present invention comprehensively monitors and analyzes various data obtained during the artifact adjustment and evaluation process, providing a solid foundation for the continuous improvement of artifact suppression methods and dynamic intervention strategies. By constructing a closed - loop optimization mechanism for the entire process of artifact processing, the technical solution can achieve dynamic regulation of the entire process from artifact detection to quality restoration, significantly improving the overall effect of face image quality assessment. Compared with traditional methods, this solution not only improves the reliability and accuracy of face recognition in dynamic light source environments but also provides more intelligent and efficient technical support for practical scenarios such as advertising placement, thus meeting the needs of commercial applications while improving system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0063] Figure 1 It is a flowchart of the face image quality assessment method, system, and computer - readable storage medium of the present invention.

[0064] Figure 2 It is a module diagram of the face image quality assessment method, system, and computer - readable storage medium of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0065] Now, the exemplary embodiments will be described more comprehensively with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that this disclosure will be more thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0066] The present invention provides as Figure 1The face image quality assessment method shown below specifically includes the following steps:

[0067] Under a dynamic light source environment, perform artifact effect detection on the face images captured by the camera in real time. By analyzing the brightness distribution uniformity, abnormal gradient direction, and texture feature changes of the image, determine whether there is an artifact effect;

[0068] The artifact effect refers to the phenomenon in face images where, due to dynamic light sources (such as neon lights or reflected light), strong reflections occur on certain areas of the face (such as the forehead or nose bridge), and through light propagation, adjacent areas (such as around the eyes) are affected, resulting in unnatural brightness transitions, abnormal textures, or detail loss in local areas of the image. This effect can mask the true features of the face image and affect its reliability in quality assessment and recognition. To detect the artifact effect, it is possible to comprehensively analyze the overall brightness distribution of the image, the edge features of local areas, and the integrity of the texture, identify the interference features of light reflection on different areas of the image, and determine whether there is an artifact effect. This detection process can adapt to changes in dynamic light sources in real time and provide support for subsequent processing.

[0069] By analyzing the brightness distribution uniformity, edge direction features, and texture changes of the image, the main interferences of the artifact effect on the image can be comprehensively captured. The analysis of the brightness distribution can identify over-bright or over-dark areas caused by light reflection. The analysis of edge features can detect the damage of the artifact effect to the normal texture direction. The analysis of texture integrity can reveal detail loss or patchy texture interference caused by artifacts. This multi-dimensional comprehensive analysis can effectively distinguish the differences between the artifact effect and normal image features, ensuring the accuracy and comprehensiveness of the detection results.

[0070] Detecting the artifact effect is a key step in this technical solution. Its purpose is to accurately identify the quality problems caused by dynamic light reflection and provide basic support for subsequent area division and intervention. Under a dynamic light source environment, the artifact effect will cause abnormalities and information loss of local features, and it is difficult for existing technologies to effectively identify this interference. If the artifact effect is not detected first, the subsequent area division and dynamic intervention mechanism will lack pertinence, which may lead to distorted quality assessment results. This solution provides an accurate basis for subsequent artifact classification, dynamic intervention, and suppression mechanisms through real-time detection and accurate judgment of the artifact effect, solving the problem of insufficient handling of the artifact effect in existing technologies.

[0071] In the case of determining that there is an artifact effect, evenly divide the face images captured by the camera in real time into several regions through the facial key point detection method;

[0072] In the case of determining the existence of artifact effects, the face images captured by the camera in real time are evenly divided into several regions through the facial key-point detection method, which is to more accurately evaluate the influence of artifact effects on different parts of the face. The facial key-point detection method can quickly identify the positions and contours of key facial regions such as eyes, nose bridge, forehead, mouth, and chin, and based on this, divide the face image into several functional regions, such as the forehead, periorbital area, nose bridge, etc. At the same time, in order to capture the cross-region propagation phenomenon of artifact effects, boundary buffer zones can also be defined at adjacent regions (such as the junction of the forehead and eyes) to achieve fine-grained division between regions. This division method can be implemented by software. Using a facial key-point localization algorithm (such as a convolutional neural network model based on deep learning) to detect the feature points of the face, and taking these key points as references, methods such as regional grid division or polygon segmentation are used to complete the uniform division of the image. This regional division method can not only achieve the localized management of artifacts, but also avoid the quality assessment errors that may be caused by global analysis, ensuring that the artifact effects of each region can be independently evaluated. Combining our technical solution, such a division method can provide precise regional support for subsequent artifact quality assessment and dynamic intervention mechanisms, solve the problem of evaluation confusion caused by the influence of artifact effects on local features in the prior art, and ultimately improve the accuracy and robustness of face image quality assessment in a dynamic light source environment.

[0073] Obtain the artifact quality assessment information of each region of the face image in real time, and analyze it after obtaining, evaluate the degree of quality impact of the artifacts on each region, and divide each region into high-impact regions, medium-impact regions, and low-impact regions according to the evaluation results;

[0074] In this embodiment, obtaining the artifact quality assessment information of each region of the face image in real time, and analyzing it after obtaining, evaluating the degree of quality impact of the artifacts on each region, and dividing each region into high-impact regions, medium-impact regions, and low-impact regions according to the evaluation results. What are the specific steps?

[0075] Obtain the artifact quality assessment information of each region of the face image in real time, and perform preprocessing after obtaining;

[0076] Real-time acquisition of artifact quality assessment information can be achieved by dynamically analyzing the face images captured by the camera. Specifically, first, the image data transmitted in real-time by the camera is subjected to region segmentation, dividing the entire image into multiple regions. Then, through image processing algorithms, the luminance distribution characteristic information and texture structure characteristic information are extracted region by region. The acquisition of the luminance distribution characteristic information can be completed by calculating the statistical data of the luminance values of the pixels in each region (such as the mean and variance), and the acquisition of the texture structure characteristic information can be achieved by applying a high-pass filter to extract the high-frequency detail components of the region and combining with the gradient calculation algorithm to analyze the texture direction characteristics and consistency of the region. The entire acquisition process is processed frame by frame to achieve real-time update of the face image quality assessment information in a dynamic light source environment.

[0077] Preprocessing is carried out to ensure the accuracy and consistency of the assessment information, thus laying a foundation for subsequent analysis. The preprocessing mainly includes three steps: noise filtering, brightness normalization, and boundary smoothing. Noise filtering removes the noise interference that may be introduced by the sensor or the external environment in the image through median filtering or Gaussian filtering; brightness normalization performs linear stretching or contrast stretching on the brightness values of each region to eliminate the overall brightness difference caused by the change in camera lighting; boundary smoothing corrects the abrupt characteristics at the region edges through convolution operations, making the cross-region characteristic analysis more accurate. These preprocessing processes are implemented by software and can be gradually completed in a pipeline manner in the real-time image processing framework to ensure that the information extracted and analyzed subsequently is more reliable.

[0078] Extract the luminance distribution characteristic information and texture structure characteristic information in the artifact quality assessment information of each region of the preprocessed face image, and perform analysis after extraction to generate the luminance balance deviation index and the detail texture damage index for each region respectively;

[0079] Extracting the brightness distribution characteristic information and texture structure characteristic information of each region of the preprocessed face image can be achieved by performing pixel-by-pixel analysis on the regional image data through image processing algorithms. Specifically, the preprocessed face image is first divided into multiple functional regions (such as forehead, eyes, bridge of nose, etc.) according to the regional division rules, and the pixel data of each region is extracted. For the brightness distribution characteristic information, the brightness value of the pixels in each region is calculated pixel by pixel, and the average value, variance, and brightness difference between the boundary pixel and the neighboring boundary pixel are counted to reflect the brightness distribution state and boundary transition characteristics of the region; and for the texture structure characteristic information, the gradient vector of each pixel is calculated through the gradient calculation algorithm (such as Sobel operator), and the gradient amplitude, direction, and gradient change rate in the time series are further obtained, and the average value and distribution variance are counted in the region to reflect the significance and directional consistency of the texture. These operations are completed through image segmentation, pixel-by-pixel analysis and regional statistics, which can extract the comprehensive characteristics of brightness and texture in real time, and provide comprehensive characteristic data support for artifact impact assessment.

[0080] An artifact impact assessment model is constructed for the brightness balance deviation index and detail texture destruction index of each generated area, and the artifact impact coefficient of each area is generated. After the generation, the model is analyzed to evaluate the impact of the artifact on the quality of each area. According to the evaluation results, each area is divided into high-impact area, medium-impact area and low-impact area.

[0081] In this embodiment, the logic for obtaining the brightness balance deviation index and detail texture destruction index of each region is as follows:

[0082] The brightness distribution characteristic information in the artifact quality assessment information of each area of ​​the preprocessed face image is extracted, including the average and variance of the brightness values ​​of all pixels in each area of ​​the face image at different times over a period of time, and the absolute average difference between the brightness values ​​of the edge pixels and the neighboring edge pixels of each area, and calibrated as , and , Indicates that within a period of time Moment face image The average brightness value of all pixels in the area Indicates that within a period of time Moment face image The variance of the brightness values ​​of all pixels in the region is Indicates that within a period of time Moment The absolute average difference between the brightness values ​​of the edge pixels in the region and the edge pixels in the neighborhood, , , and are all positive integers;

[0083] The average value of the brightness values of all pixels in each region of the face image at different times within a period: First, in each frame of the image, calculate the arithmetic mean of the brightness values of all pixels for each region (i.e., the brightness mean), and then take the average of the brightness means at different times in the time series to obtain the brightness mean of the region over the entire period, reflecting the dynamic characteristics of the overall brightness level of the region.

[0084] The variance of the brightness values of all pixels in each region of the face image at different times within a period: For the brightness data of each region in each frame of the image, calculate the variance of the pixel brightness values within the region to measure the severity of the brightness change in the region, and then perform time averaging on the variance data at different times to generate the time characteristics of the brightness variance of the region. This data reflects the degree of damage to the smoothness of the local region brightness distribution caused by artifacts.

[0085] The absolute average difference between the brightness values of the edge pixels in each region and the neighboring edge pixels at different times within a period: Use the boundary detection algorithm to identify the boundary pixels between the current region and the neighboring region, calculate the absolute difference between the brightness value of the boundary pixel in the current region and the brightness value of the boundary pixel in the neighboring region for each pixel, and take the average of the differences of all boundary pixels to obtain the absolute average difference. This data dynamically reflects the cross-region brightness change characteristics caused by artifacts at the region boundary.

[0086] Calculate the brightness balance deviation index for each region. The specific calculation formula is as follows:

[0087]

[0088] In the formula, is the brightness balance deviation index of the th region;

[0089] Brightness balance deviation index The calculation formula comprehensively evaluates the influence of artifacts on the brightness distribution of each region and the boundary brightness consistency through multi-stage operation steps, thereby realizing the comprehensive quantification of artifact interference. Taking the average of the calculation results at multiple times within a period in the formula can effectively eliminate the noise interference caused by instantaneous illumination fluctuations in a single-frame image and ensure the stability of the evaluation results. Subsequently, the part in the formula evaluates the comprehensive influence of artifacts on the brightness distribution within the region and at the region boundary by normalizing the product of the brightness difference and the edge brightness deviation with the average brightness , while avoiding the amplification effect of extreme cases of too high or too low brightness on the results. And The item magnifies the brightness deviation of the region boundary in logarithmic form The contribution to the evaluation result further highlights the interference effect of artifacts in cross - region propagation. Through the synergistic effect of these operation steps, the formula can comprehensively capture the impact of artifacts on regional brightness uniformity and boundary consistency, providing scientific and accurate quantitative results.

[0090] The brightness balance deviation index of the th region directly reflects the strength of the impact of artifacts on the quality of this region. The larger the index value, the more serious the interference of artifacts on the brightness distribution of this region. Specifically, integrates the dynamic characteristics of the brightness fluctuation degree within the region ( ), the brightness discontinuity at the region edge ( ), and the overall brightness level ( ). If has a large value, it indicates that artifacts have caused uneven changes in brightness within the region; if has a high value, it means that artifacts have caused cross - region light propagation interference at the region boundary; while a generally low will amplify the relative significance of the above - mentioned interference effects. By integrating these factors into an index value, can quantify the degree of damage of artifacts to brightness uniformity and boundary consistency. The regions with higher index values are evaluated as being more interfered with by artifacts in terms of quality and may require stronger intervention measures.

[0091] Extract the texture structure characteristic information in the artifact quality evaluation information of each region of the pre - processed face image, specifically including the average change rate, amplitude, and direction distribution variance of pixel gradients in each region of the face image at different times within a period of time, and calibrate them as , and , represents the average change rate of pixel gradients in the th region of the face image at the th time within a period of time, represents the amplitude of pixel gradients in the th region of the face image at the th time within a period of time, represents the direction distribution variance of pixel gradients in the th region of the face image at the th time within a period of time;

[0092] The acquisition and interpretation of these three types of quantitative data are as follows: Gradient is a vector that describes the rate of change of pixel brightness in an image and is usually used to measure the brightness change in the image. The average rate of change of pixel gradients is obtained by calculating the brightness changes of all pixel points in the same region frame by frame and is used to reflect the detailed change characteristics caused by artifacts. Specifically, within a period of time, artifacts will cause the gradient changes within the region to become drastic or uneven, thus affecting the detailed characteristics. The magnitude of the pixel gradient represents the intensity of pixel brightness change. The presence of artifacts usually reduces the gradient magnitude, blurring the image edges and making the texture less prominent. This data is obtained by analyzing the strength of brightness changes pixel by pixel within the region and averaging over the entire region. The variance of the direction distribution of pixel gradients describes the direction consistency of pixel brightness changes within the region and evaluates the damage to texture directionality caused by artifacts by analyzing the degree of dispersion of gradient directions. Artifacts will cause disordered changes in gradient directions, thus significantly increasing the dispersion of the direction distribution. These three types of data comprehensively reflect the comprehensive impact of artifacts on regional detailed textures from three perspectives: the rate of change, the intensity of change, and the direction consistency.

[0093] Calculate the detailed texture damage index for each region. The specific calculation formula is as follows:

[0094]

[0095] In the formula, is the detailed texture damage index of the th region.

[0096] This formula comprehensively quantifies the damage degree of artifacts to the detailed textures of each region through multiple calculation steps, with a reasonable design and strong pertinence. First, the formula performs time averaging, which can smooth the noise and sudden interferences in a single-frame image within a period of time to ensure the stability of the evaluation results. The first part of the formula comprehensively considers the detailed gradient change rate ( ) and the gradient magnitude ( ) within the region. The product of the two quantifies the dynamic changes and edge saliency of regional details. The presence of artifacts will reduce these values, while the variance of the gradient direction distribution ( ) in the denominator makes the contribution of the damage to direction consistency caused by artifacts more prominent by magnifying the impact of disordered direction changes. The second part of the formula further evaluates the relationship between the degree of disorder in direction distribution and the rate of detailed changes. The square root operation is used to smooth the extreme values that may occur when artifacts are severe to avoid distortion of the evaluation results. The overall formula design comprehensively reflects the damage degree of artifacts to regional detailed textures from three dimensions: detailed changes, texture intensity, and direction consistency. At the same time, through the averaging calculation in the time dimension, it enhances the adaptability to dynamic light source scenarios and the accuracy of the evaluation results.

[0097] The Detail texture damage index of an area The size directly reflects the degree of damage to the detail texture of the area by the artifacts, and has a positive correlation with evaluating the degree of influence of the artifacts on the area quality. Specifically, the larger the value, the more serious the influence of the detail texture of the area by the artifacts, including problems such as abnormal detail change rate, reduced texture edge saliency, and disrupted gradient direction consistency. A higher index value indicates that the disorder of the texture direction caused by the artifacts is significant, and the texture details and edge characteristics within the area are severely damaged, directly affecting the image quality and the reliability of feature extraction in this area. Through this index, the degree of interference of the artifacts on the texture quality of each area can be quantified, and areas with higher indices require stronger intervention measures to repair their quality problems.

[0098] In this embodiment, for the brightness balance deviation index and the detail texture damage index of each generated area, an artifact influence evaluation model is constructed, and the artifact influence coefficient of each area is generated by weighted summation, and the generated artifact influence coefficient of each area is compared with the pre-set artifact influence coefficient threshold interval to evaluate the degree of influence of the artifacts on the quality of each area according to the comparison result, and each area is divided into a high-influence area, a medium-influence area, and a low-influence area according to the evaluation result. The specific comparison analysis and division are as follows:

[0099] If , the degree of influence of the artifacts on the quality of this area is low, and this area is divided into a low-influence area;

[0100] This situation means that the degree of influence of the artifacts on the quality of this area can be ignored, and the brightness distribution and texture characteristics within the area basically remain stable, without significant abnormalities or damage. Due to the weak artifact interference, the image quality of this area is close to the original state and does not require further repair or intervention. Such areas usually do not have an adverse impact on subsequent image analysis (such as face recognition or feature extraction) and can be directly used for related tasks, thereby improving the processing efficiency of the system and reducing the waste of computing resources.

[0101] If , the degree of influence of the artifacts on the quality of this area is medium, and this area is divided into a medium-influence area;

[0102] This situation indicates that the impact of artifacts on the quality of this area is relatively obvious, but it has not reached the level of severe damage. There may be certain brightness fluctuations or local damage to texture characteristics within the area, such as blurred edges or slightly disordered direction consistency, which may interfere with subsequent image analysis. For such areas, the system needs to take appropriate intervention measures according to the specific situation of the artifact impact, such as adjusting brightness or enhancing texture, to restore the image quality of the area as much as possible and ensure that it meets the requirements of subsequent analysis tasks.

[0103] If , and the impact of artifacts on the quality of this area is at a high level, then this area is classified as a high-impact area.

[0104] This situation means that the impact of artifacts on the quality of this area is very serious. The brightness distribution and texture characteristics may have been significantly damaged, resulting in a substantial decline in the image quality of the area. For example, there may be overexposure or underexposure of brightness, as well as texture loss or complete disorder of direction consistency. If such areas are not effectively repaired, it will significantly affect the accuracy of subsequent face feature extraction or recognition. To reduce the interference of artifacts, the system must take strong intervention measures for high-impact areas, such as brightness correction, texture enhancement, or artifact suppression algorithms, to maximize the restoration of image quality and ensure the normal progress of subsequent tasks.

[0105] The specific implementation of "constructing an artifact impact assessment model for the brightness balance deviation index and detail texture damage index of each generated area, and generating the artifact impact coefficient of each area through weighted summation" is as follows: First, take the brightness balance deviation index and the detail texture damage index of each area as input variables, and assign weight coefficients based on their importance in artifact impact assessment to construct a weighted assessment model. The selection of weight coefficients can be set according to the different contributions of the two types of indicators to artifact impact. For example, if the damage to brightness balance by artifacts is more significant in a certain scenario, a higher weight is assigned to the brightness balance deviation index, and a lower weight is assigned to the detail texture damage index, and the sum of weights is always 1 (such as and , satisfying ). In the model, calculate the artifact impact coefficient ( ) of the th area through the formula ), and respectively reflect the contribution ratios of the two indicators to artifact impact. This model synthesizes brightness and texture characteristics into a unified quantitative index. By adjusting the weight coefficients, it can flexibly adapt to the main characteristics of artifact interference in different scenarios, making the evaluation results more targeted and accurate, and providing a scientific basis for area division and intervention measures.

[0106] The pre-set artifact influence coefficient threshold range can be automatically determined by analyzing a large number of sample data and combining statistical methods. The specific implementation method is as follows: First, collect a face image dataset under various typical dynamic light source environments, covering regions with different degrees of artifact influence (such as low-influence, medium-influence, and high-influence regions). Then, calculate the brightness balance deviation index and detail texture damage index for each sample region through image processing algorithms to further generate the artifact influence coefficient. The values of all sample regions are labeled and classified according to the degree of influence (such as manually labeled or assisted by other verified quality assessment models for classification), and the value distribution of each classification is statistically analyzed. By calculating the mean and standard deviation of the values of each classification, the classification boundaries are determined. For example, the upper limit ( ) of the low-influence region can be taken as the maximum value of the values in the low-influence region, and the lower limit ( ) of the high-influence region can be taken as the minimum value of the values in the high-influence region, or the interval division can be further refined by setting quantiles (such as the 25th quantile, 75th quantile). Finally, these thresholds are stored in the system as parameters of the model to form a dynamically adjustable artifact influence coefficient threshold range, enabling the model to have stronger adaptability and classification accuracy under different environments or data distributions.

[0107] According to the division results of each region of the face image, a dynamic intervention mechanism is constructed, and corresponding intervention measures are taken for the high-influence region, medium-influence region, and low-influence region respectively;

[0108] In this embodiment, according to the division results of each region of the face image, a dynamic intervention mechanism is constructed, specifically: according to the division results of the high-influence region, medium-influence region, and low-influence region, different intervention measure parameters are set respectively to form a dynamic intervention mechanism; this dynamic intervention mechanism is based on the artifact influence coefficient of each region and automatically determines the adjustment method and amplitude of brightness adjustment, texture enhancement, and edge smoothing through pre-set rules;

[0109] The construction of a dynamic intervention mechanism can be achieved by combining the rule engine and image processing algorithms in a software system. The specific methods are as follows: First, based on the calculation results of the artifact influence coefficients for each region of the face image, the rule engine presets intervention strategies for the high-influence region, medium-influence region, and low-influence region. These strategies include the specific adjustment methods and amplitudes for brightness adjustment, texture enhancement, and edge smoothing. The rule engine dynamically matches intervention measures according to the range of the artifact influence coefficients. For example, it assigns enhanced parameters to the high-influence region to strengthen the adjustment of brightness uniformity, enhance texture characteristics, and smooth the edge transition; it assigns moderately optimized parameters to the medium-influence region to moderately adjust the brightness and texture; it assigns monitoring parameters to the low-influence region to only monitor the quality status. In specific implementation, the brightness normalization algorithm in the image processing module is called to adjust the brightness distribution consistency within the region, the texture enhancement algorithm (such as local contrast enhancement or high-frequency detail amplification) is used to optimize the texture characteristics within the region, and the edge detection and smoothing algorithm are combined to eliminate the boundary discontinuity or mutation caused by artifacts. The purpose of this is to flexibly adjust the intervention measures according to the degree of influence of artifacts on the quality of each region, ensure the maximization of resource utilization efficiency, and maintain the regional stability in a dynamic lighting environment while ensuring the image quality. Ultimately, it provides reliable image data support for face recognition and analysis tasks.

[0110] Corresponding intervention measures are taken for the high-influence region, medium-influence region, and low-influence region respectively. Specifically: within the high-influence region, the artifact suppression enhancement parameters in the dynamic intervention mechanism are used to significantly adjust the brightness distribution, strengthen texture enhancement, and optimize edge smoothing; within the medium-influence region, the artifact moderate adjustment parameters in the dynamic intervention mechanism are used to moderately optimize the brightness distribution and texture characteristics; within the low-influence region, the artifact stable parameters in the dynamic intervention mechanism are maintained without adjustment, and only the regional status is monitored to maintain quality stability.

[0111] Corresponding intervention measures are implemented for high - impact areas, medium - impact areas, and low - impact areas respectively, which can be achieved through the combination of dynamic parameter control and image - processing algorithms. Specifically, the software system first matches the intervention strategy of each area with its artifact influence degree based on the artifact suppression parameters, moderate adjustment parameters, or stability parameters allocated by the dynamic intervention mechanism. In high - impact areas, the brightness distribution is significantly adjusted through the brightness adjustment algorithm. For example, overexposure or underexposure phenomena are eliminated through histogram equalization or brightness stretching algorithms; high - frequency details are enhanced and texture characteristics are restored through texture enhancement algorithms (such as local contrast enhancement or Laplacian operator); at the same time, the boundary mutations caused by artifacts are optimized by combining edge - smoothing algorithms (such as Gaussian filtering or convolution operations) to ensure a smooth transition of edge characteristics. In medium - impact areas, a moderate optimization strategy is adopted to make small adjustments to the brightness distribution and texture characteristics, such as moderate brightness mean normalization and contrast enhancement, to reduce the influence of artifacts without excessive intervention and ensure resource utilization efficiency. In low - impact areas, the artifact stability parameters are used to maintain the current state, and only the brightness, texture, and edge characteristics of the area are dynamically monitored through real - time monitoring algorithms to ensure the stability of the area quality without increasing the computational burden. The purpose of implementing hierarchical intervention is to precisely adjust the intervention intensity according to the artifact influence degree, maximize the restoration of the image quality in high - impact areas, and avoid unnecessary operations on medium - and low - impact areas, thereby balancing system performance and processing efficiency and ensuring the balance and stability of the overall image quality.

[0112] During the process of the dynamic intervention mechanism intervening in each area, the intervention feedback information of each area is obtained in real - time, and after obtaining it, an analysis is carried out to evaluate whether the intervention effect of the dynamic intervention mechanism in each area can meet the expectations, and the dynamic intervention mechanism is optimized according to the evaluation results;

[0113] In this embodiment, during the process of the dynamic intervention mechanism intervening in each area, the intervention feedback information of each area is obtained in real - time, and after obtaining it, an analysis is carried out to evaluate whether the intervention effect of the dynamic intervention mechanism in each area can meet the expectations, and the dynamic intervention mechanism is optimized according to the evaluation results, which specifically includes the following steps:

[0114] During the process of the dynamic intervention mechanism intervening in each area, the intervention feedback information of each area is obtained in real - time and pre - processed after obtaining it;

[0115] Real-time acquisition of intervention feedback information for each region can be achieved by dynamically monitoring the real-time adjustment effect of the intervention mechanism on the image. Specifically, during the process of dynamically intervening in each region, the system continuously captures changes in the brightness and texture characteristics of each region based on the image processing module. By analyzing the image data frame by frame after intervention, the pixel brightness values and gradient information of each region are extracted, including key data such as the average brightness of the region, the brightness distribution characteristics, as well as the texture gradient amplitude, gradient direction, and direction distribution variance. These data are analyzed in real time through algorithms and attributed to each region, forming comprehensive feedback information including the adjustment effects of brightness and texture. In addition, by capturing multi-frame feedback data through time series processing methods, a continuous evaluation dataset of the dynamic intervention effect is established to ensure that the feedback information can comprehensively reflect the actual impact of the intervention on the image.

[0116] The purpose of preprocessing is to improve the accuracy and stability of the feedback information and provide a reliable data basis for subsequent analysis. After the feedback information is obtained, a series of preprocessing operations need to be performed on it, mainly including denoising, normalization, and outlier detection. Gaussian filtering or median filtering methods can be used for denoising to eliminate random errors that may be introduced due to sensor noise or external environmental interference in the image data; normalization processing unifies the brightness values and gradient amplitudes of each region to the same magnitude through linear stretching or contrast adjustment, facilitating subsequent analysis and comparison; outlier detection eliminates possible incorrect data points through statistical methods (such as mean and standard deviation range judgment) to ensure the overall credibility of the feedback data. These preprocessing operations are gradually completed through a software pipeline, providing high-quality input data for the subsequent generation of the brightness adjustment deviation index and the texture recovery consistency index.

[0117] Extract the brightness adjustment feedback information and texture optimization feedback information from the intervention feedback information of each region after preprocessing, and analyze them after acquisition to generate the brightness adjustment deviation index and texture recovery consistency index for each region respectively;

[0118] Extracting the brightness adjustment feedback information and texture optimization feedback information from the intervention feedback information of each preprocessed region can be achieved through a multi-dimensional data analysis method based on regional characteristics. Specifically, first, the preprocessed intervention feedback information is divided according to regions, and the brightness and texture-related characteristic data within each region are extracted respectively. For the brightness adjustment feedback information, by analyzing the statistical characteristics of the pixel brightness values within the region, data including the brightness mean, brightness variance, etc. are extracted to measure the adjustment effect of the overall brightness of the region and the consistency of the brightness distribution. For the texture optimization feedback information, the gradient amplitude and direction distribution characteristics of each region are calculated through an image gradient algorithm, and the texture intensity and direction consistency features of the region after intervention are extracted, including key data such as the average value of the gradient amplitude and the variance of the gradient direction. These data form the brightness adjustment feedback information and texture optimization feedback information of each region through pixel-by-pixel calculation and regional summary, and are uniformly managed through a standardized storage structure, providing detailed regional characteristic data support for subsequent index calculation and intervention effect evaluation.

[0119] Construct an intervention effect evaluation model for the brightness adjustment deviation index and texture restoration consistency index of each generated region, generate the intervention coefficient of each region, compare the generated intervention coefficient of each region with the pre-set intervention coefficient threshold of each region, evaluate whether the intervention effect of this dynamic intervention mechanism in each region can meet the expectations according to the comparison result, and optimize the dynamic intervention mechanism according to the evaluation result.

[0120] The pre-set intervention coefficient threshold of each region can be achieved through statistical analysis of historical data. The specific method is to collect face image samples before and after intervention in multiple scenarios, calculate the brightness adjustment deviation index and texture restoration consistency index of each region, and generate the intervention coefficient. Classify and label the intervention coefficients according to the quality of the intervention effect of the samples, and calculate the mean, median and standard deviation of the distribution of the intervention coefficients of various intervention effects through statistical analysis. Use the median or mean of the distribution of highly efficient intervention samples as the ideal range, and combine quantiles or ± standard deviations to set the threshold range to ensure rationality and adaptability. At the same time, the threshold setting can be dynamically adjusted according to the requirements of different scenarios or optimized through machine learning.

[0121] In this embodiment, the acquisition logic of the brightness adjustment deviation index and texture restoration consistency index of each region is as follows:

[0122] Extract the brightness adjustment feedback information from the intervention feedback information of each preprocessed region, specifically including the average value, variance of the pixel brightness values of each region at different times within a period of time after intervention, and the pre-set average value of the pixel brightness values of each region after intervention, and calibrate them respectively as 、 and , Indicates within a period of time after the intervention At the th moment, the average pixel brightness value of the th region, At the th moment, the variance of the pixel brightness value of the th region, Indicates the average pixel brightness value of the th region preset after the intervention, , and are all positive integers;

[0123] Within a period of time after the intervention, the average value, variance, and preset target brightness value of the pixel brightness of each region at different moments can all be dynamically obtained through image processing algorithms and system preset parameters. First, the system captures the image data after the intervention in real time through a camera, and divides the face image into multiple independent functional regions (such as the forehead, eyes, nose bridge, etc.) according to the region division rules, and extracts the pixel brightness value data frame by frame for each region. For the average value of the pixel brightness value, it is obtained by taking the arithmetic mean of all pixel brightness values in this region for each frame; for the variance of the pixel brightness value, it is completed by calculating the deviation degree of each pixel brightness value in the region from the brightness mean value of the region and taking the variance, which is used to reflect the uniformity of the brightness distribution. The above two types of data are collected at multiple moments within a period of time after the intervention, and the average is taken according to the time series to dynamically generate the core data reflecting the adjustment effect of the regional brightness distribution. As for the target brightness value, it is the quantitative data preset according to the scene requirements or ideal quality goals when setting the intervention mechanism, and it is an important reference for comparing and evaluating the intervention effect. This acquisition method not only ensures the dynamic real-time nature of the data but also combines the guidance of the target value, providing a scientific basis for the quantitative evaluation of the brightness adjustment effect.

[0124] Calculate the brightness adjustment deviation index for each region. The specific calculation formula is as follows:

[0125]

[0126] In the formula, is the brightness adjustment deviation index of the th region;

[0127] The design of calculating the brightness adjustment deviation index in the above way is mainly to comprehensively and dynamically evaluate the effect of the dynamic intervention mechanism on the adjustment of the regional brightness distribution. By obtaining the brightness data at multiple moments within a period of time after the intervention and taking the average value, it is possible to effectively eliminate the fluctuations and noise interferences that may exist in the instantaneous data, ensuring that the calculation results are more stable and reliable. The two parts in the formula - the brightness mean deviation and the brightness distribution variance - respectively quantify the overall deviation degree and the distribution uniformity of the brightness adjustment. This design can not only capture the global adjustment effect but also reflect the detailed characteristics of the local brightness inconsistency. At the same time, by normalizing the brightness variance and combining it with the brightness mean deviation, it can more intuitively show whether the consistency of the brightness distribution after the intervention has been improved. Especially in the dynamic light source scenario, this calculation method can more comprehensively reflect the intervention effect, providing accurate and quantifiable basis for subsequent optimization.

[0128] The brightness adjustment deviation index of the nth area directly reflects whether the brightness adjustment effect of the dynamic intervention mechanism in this area meets the expectation. The larger the index, the higher the degree of deviation of the brightness distribution after the intervention from the target value, and the worse the adjustment effect; the smaller the index, the closer the brightness adjustment is to the target value, and the more ideal the effect. Specifically, the first part of the quantifies the deviation between the brightness mean after the intervention and the target value. If the deviation value is large, it indicates that the overall brightness adjustment has not reached the ideal level; the second part reflects the uniformity of the brightness distribution after the intervention through the brightness variance. If the distribution is uneven, it indicates that there are still deficiencies in the local brightness adjustment. By comparing the index value with the pre-set evaluation threshold range, it is possible to clarify whether the effect of the dynamic intervention mechanism in this area exceeds the expectation, meets the expectation, or fails to meet the expectation, providing a scientific basis for further optimizing the intervention strategy.

[0129] Extract the texture optimization feedback information in the intervention feedback information of each pre-processed area, specifically including the average amplitude, direction distribution variance of the pixel gradients in each area at different moments within a period of time after the intervention, and the pre-set average amplitude of the pixel gradients in each area after the intervention, and respectively label them as , and , represents the average amplitude of the pixel gradients in the nth area at the mth moment within a period of time after the intervention, represents the direction distribution variance of the pixel gradients in the nth area at the mth moment within a period of time after the intervention, represents the pre-set average amplitude of the pixel gradients in the nth area after the intervention;

[0130] The real-time image processing algorithm can be used to obtain the average amplitude of the pixel gradients in each region at different times, the directional distribution variance, and the pre-set average amplitude of the pixel gradients in each region after the intervention. First, the gradient vector of each pixel is calculated for each pixel point in each region using the gradient calculation algorithm (such as the Sobel operator or the Scharr operator) through the image captured by the camera in real time after the intervention, and the gradient amplitude and gradient direction are extracted from it. For the average amplitude of the gradient, the gradient amplitude of all pixels in the region is averaged to reflect the texture intensity level of the region; for the directional distribution variance, the variance of the gradient direction of all pixels in the region is calculated to quantify the consistency of the texture direction of the region after the intervention; and the pre-set average amplitude of the gradient is based on the intervention goal. The ideal gradient amplitude is set under the condition of artifact elimination, which can usually be determined by analyzing the artifact-free image data set or expert experience. These data comprehensively describe the texture characteristics of each region after the intervention from three dimensions: brightness change, texture saliency, and directional consistency, providing an accurate quantitative basis for evaluating the intervention effect.

[0131] Calculate the texture restoration consistency index of each area. The specific calculation formula is as follows:

[0132]

[0133] In the formula, For the Texture restoration consistency index for each region.

[0134] The calculation result of texture restoration consistency index is to take the average value of data at multiple moments in a period of time as the purpose of comprehensively and stably reflecting the overall restoration effect of dynamic intervention on regional texture characteristics. The texture characteristics of each region after intervention may gradually stabilize over time or show short-term fluctuations in different frames. By averaging the gradient amplitude, directional distribution variance and other data at multiple moments, the interference of instantaneous fluctuations on the results can be effectively smoothed, and the evaluation error caused by accidental or noise data at a single moment can be avoided. In addition, this time dimension processing can more truly reflect the overall trend and cumulative impact of the intervention effect, so that the calculation results can not only evaluate the immediate state of texture restoration after intervention, but also reflect the long-term effect, providing a more reliable and scientific basis for further optimizing the intervention mechanism.

[0135] No. Texture restoration consistency index of the region The size of directly reflects the degree of deviation between the intervention effect of the dynamic intervention mechanism in this area and the expected goal. Specifically, The smaller the value, the closer the texture characteristics (such as gradient magnitude and direction consistency) of the intervened area are to the preset target value, indicating that the effect of the dynamic intervention mechanism in this area is closer to the expectation; conversely, the larger the value, the more insufficient the texture intensity or the higher the disorder of the direction distribution in the intervened area, indicating that the intervention effect fails to fully restore the area quality and there is still much room for improvement. Therefore, by evaluating the size of, the success degree of texture restoration after intervention can be quantified, and based on this, it can be judged whether the dynamic intervention mechanism has achieved the set quality target in each area, providing a clear direction for optimizing the intervention strategy.

[0136] In this embodiment, for the brightness adjustment deviation index and texture restoration consistency index of each generated area, an intervention effect evaluation model is constructed, and the intervention coefficient of each area is generated by weighted summation, and the generated intervention coefficient of each area is compared with the intervention coefficient threshold of each preset area, and according to the comparison result, it is evaluated whether the intervention effect of the dynamic intervention mechanism in each area can meet the expectation, and the dynamic intervention mechanism is optimized according to the evaluation result. The specific comparison and analysis are as follows:

[0137] If , the intervention effect of the dynamic intervention mechanism in this area can meet the expectation and there is no need to optimize the dynamic intervention mechanism;

[0138] This situation means that the intervention effect of the dynamic intervention mechanism in this area has met the expectation, and the brightness distribution and texture characteristics of the area have been successfully restored to the ideal state or close to the target level. The brightness adjustment deviation index and the texture restoration consistency index are both within the acceptable range, indicating that the influence of artifacts after intervention has been effectively eliminated or significantly reduced and will not cause negative interference to the image quality of the area. In this case, there is no need to further adjust the dynamic intervention mechanism, and the current parameter settings can be maintained, which not only saves computing resources but also ensures the system operation efficiency, providing high-quality data support for subsequent image analysis.

[0139] If , the intervention effect of the dynamic intervention mechanism in this area cannot meet the expectation, and the dynamic intervention mechanism needs to be optimized, specifically including: increasing the brightness adjustment intensity of this area to reduce the brightness distribution deviation; increasing the parameter weight of texture enhancement to strengthen the contrast of details and high-frequency texture characteristics; adjusting the edge smoothing algorithm to reduce the boundary mutation problem caused by artifact interference; combining the real-time intervention feedback data of this area to dynamically adjust the intervention parameters and rules and optimize the execution process of the intervention mechanism.

[0140] This situation indicates that the intervention effect of the dynamic intervention mechanism in this area has not reached the expected level, and there are still significant deviations in the brightness and texture characteristics of the area. For example, the brightness adjustment is insufficient or the consistency of the texture direction is not restored, which may lead to significant interference from artifacts. Such results may interfere with subsequent image processing steps (such as feature extraction or face recognition), reducing the accuracy and reliability of the system. At this time, it is necessary to optimize the parameters and rules of the dynamic intervention mechanism, such as strengthening brightness adjustment, increasing texture enhancement intensity, or adjusting the edge smoothing algorithm, to further improve the image quality of this area and ensure that the overall effect meets the expectations.

[0141] These optimization measures can be achieved by combining image processing algorithms and dynamic parameter regulation, as follows: First, through the brightness adjustment module in the software system, combined with the region brightness distribution data obtained in real time, dynamically adjust the weights and parameters of the brightness equalization algorithm. For example, adopt adaptive histogram equalization (CLAHE) or brightness normalization method to significantly enhance the contrast of the region brightness and reduce the brightness deviation, thereby improving the brightness consistency after intervention. Second, texture enhancement optimization can be achieved through local contrast enhancement or algorithms based on high-frequency detail extraction (such as Laplacian enhancement or wavelet transform). By increasing the significance of high-frequency details, strengthen the texture characteristics within the region, and at the same time adjust the enhancement parameter weights to make the texture restoration more significant and closer to the target effect. Edge smoothing optimization can be achieved through methods such as Gaussian filtering or bilateral filtering to smooth the boundary mutations caused by artifact interference, further improving the transition effect and visual quality of the edge region. In addition, combined with real-time intervention feedback data (such as brightness distribution deviation values, texture characteristic indicators, etc.), dynamically adjust the key parameters and algorithm execution order of the intervention mechanism through a rule engine. For example, determine the priority of brightness adjustment and texture enhancement or adjust the enhancement intensity according to the feedback data. This method can not only optimize the intervention strategy according to the region state, but also achieve dynamic optimization of the intervention effect through continuous adaptive adjustment. Through these methods, the intervention effect can be significantly improved in the end, ensuring that the impact of artifacts is minimized, while keeping the brightness and texture characteristics of the region close to the target state.

[0142] "Construct an intervention effect evaluation model for the brightness adjustment deviation index and texture restoration consistency index of each generated region, and generate an intervention coefficient for each region through weighted summation" can be achieved in the following way: First, take the brightness adjustment deviation index and texture restoration consistency index of each region as the input variables of the evaluation model, reflecting the comprehensive effect of the dynamic intervention mechanism on brightness distribution and texture restoration. Then, by setting weight coefficients and , construct a weighted summation formula , where and The value is determined by the importance of two indicators in a specific scenario. For example, in a scenario where higher requirements are placed on brightness distribution, the weight of can be increased to give a greater contribution to brightness adjustment; while in a scenario that emphasizes the restoration of texture details, the weight of can be increased to make the contribution of texture restoration to the evaluation of the intervention effect more significant. These weight coefficients can be set through experiments or according to scenario requirements to ensure the flexibility and applicability of the model. Finally, by calculating the weighted intervention coefficient for each region, comprehensively evaluate whether the intervention effect of this region meets the expectations, providing a quantitative basis for the further optimization of the dynamic intervention mechanism.

[0143] During the artifact adjustment and evaluation process, real-time monitoring and comprehensive analysis are carried out on the artifact quality evaluation information, intervention feedback information, intervention records, and image quality restoration effects obtained, continuously improving the artifact suppression method and dynamic intervention strategy, and optimizing the face image quality evaluation process.

[0144] This goal can be achieved by combining real-time data collection, analysis models, and optimization algorithms. Specifically, first, through the data collection module in the software system, real-time obtain artifact quality evaluation information (such as brightness distribution characteristics, texture characteristics, etc.), intervention feedback information (such as brightness adjustment results and texture enhancement effects, etc.), intervention records (including intervention parameter settings and rule execution situations), and image quality restoration effects (such as changes in the artifact influence coefficients of each region). These data are dynamically stored in the system's database and passed as inputs to the comprehensive analysis module. Then, through the analysis model, correlation analysis is carried out on the data. For example, using a time series analysis model to predict the trend of artifact adjustment effects, or mining the correlation between intervention parameters and artifact evaluation results based on a rule engine to identify potential optimization points in the current dynamic intervention strategy. At the same time, combined with machine learning algorithms (such as regression analysis or decision trees), train the historical intervention records to form an optimization recommendation model based on data-driven. Finally, through the optimization module, dynamically adjust the execution rules of the artifact suppression method and the parameter configuration of the intervention strategy, so that the system can enhance the accuracy of artifact adjustment and the adaptability of dynamic intervention through continuous learning and improvement. The purpose of this is to continuously optimize the overall process of face image quality evaluation through real-time monitoring and comprehensive analysis of the entire process data of artifact processing, enabling the system to more efficiently adapt to artifact interference under complex lighting conditions, and ensuring the accuracy and stability of subsequent face recognition and feature extraction.

[0145] As Figure 2 shown, the face image quality evaluation system includes an artifact effect detection module, a region division and calibration module, an artifact quality evaluation module, a dynamic intervention execution module, an intervention effect optimization module, and a comprehensive monitoring and improvement module;

[0146] An artifact effect detection module that, in a dynamic light source environment, detects the artifact effect on the face image captured by the camera in real time. By analyzing the brightness distribution uniformity, abnormal gradient direction, and texture feature changes of the image, it determines whether there is an artifact effect;

[0147] A region division and calibration module that, when it is determined that there is an artifact effect, evenly divides the face image captured by the camera in real time into several regions through the facial key point detection method;

[0148] An artifact quality evaluation module that obtains the artifact quality evaluation information of each region of the face image in real time, and analyzes it after obtaining. It evaluates the degree of quality impact of the artifact on each region, and divides each region into a high-impact region, a medium-impact region, and a low-impact region according to the evaluation results;

[0149] A dynamic intervention execution module that, according to the division results of each region of the face image, constructs a dynamic intervention mechanism, and performs corresponding intervention measures on the high-impact region, the medium-impact region, and the low-impact region respectively;

[0150] An intervention effect optimization module that, during the process of the dynamic intervention mechanism intervening in each region, obtains the intervention feedback information of each region in real time, and analyzes it after obtaining. It evaluates whether the intervention effect of the dynamic intervention mechanism in each region can meet the expectation, and optimizes the dynamic intervention mechanism according to the evaluation results;

[0151] A comprehensive monitoring and improvement module that monitors and comprehensively analyzes the artifact quality evaluation information, intervention feedback information, intervention records, and image quality restoration effect obtained during the artifact adjustment and evaluation process, continuously improves the artifact suppression method and dynamic intervention strategy, and optimizes the face image quality evaluation process.

[0152] A computer-readable storage medium that includes a stored computer program. When the computer program runs, it controls the device where the computer-readable storage medium is located to execute the face image quality evaluation method.

[0153] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0154] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can 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 processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0155] It should be understood that in various embodiments of the present application, the order of the above processes does not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0156] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0157] In several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the above-described embodiments are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be an indirect coupling or communication connection through some interfaces, devices, or units, and can be in an electrical, mechanical, or other form.

[0158] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0159] In addition, each functional unit in various embodiments of the present application may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit.

[0160] As mentioned above, the above are only specific implementation manners of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for assessing the quality of a face image, characterized in that: The specific steps include: Under dynamic light source environment, the face image captured by the camera in real time is tested for artifact effects. By analyzing the uniformity of image brightness distribution, gradient direction anomalies, and texture feature changes, it is determined whether there is an artifact effect. When it is determined that there is an artifact effect, the face image captured by the camera in real time is evenly divided into several areas through a facial key point detection method; Acquire artifact quality assessment information of each area of ​​the face image in real time, analyze it after acquisition, assess the impact of artifacts on the quality of each area, and divide each area into high-impact area, medium-impact area and low-impact area according to the assessment results; According to the division results of each area of ​​the face image, a dynamic intervention mechanism is constructed to take corresponding intervention measures for high-impact areas, medium-impact areas and low-impact areas respectively; In the process of dynamic intervention mechanism intervening in various regions, the intervention feedback information of each region is obtained in real time, and analyzed after acquisition to evaluate whether the intervention effect of the dynamic intervention mechanism in each region can meet expectations, and optimize the dynamic intervention mechanism according to the evaluation results; Real-time monitoring and comprehensive analysis of the artifact quality assessment information, intervention feedback information, intervention records and image quality restoration effects obtained during the artifact adjustment and evaluation process are carried out to continuously improve artifact suppression methods and dynamic intervention strategies and optimize the facial image quality assessment process.

2. The facial image quality assessment method according to claim 1, characterized in that: The artifact quality assessment information of each area of ​​the face image is obtained in real time, and analyzed after acquisition to assess the impact of the artifact on the quality of each area, and the areas are divided into high-impact areas, medium-impact areas and low-impact areas according to the assessment results, including: Acquire artifact quality assessment information of each area of ​​the face image in real time and perform preprocessing after acquisition; Extracting brightness distribution characteristic information and texture structure characteristic information from artifact quality assessment information of each region of the preprocessed face image, and analyzing them after extraction to generate brightness balance deviation index and detail texture destruction index of each region respectively; The logic for obtaining the brightness balance deviation index and detail texture destruction index of each area is as follows: The brightness distribution characteristic information in the artifact quality assessment information of each area of ​​the preprocessed face image is extracted, including the average and variance of the brightness values ​​of all pixels in each area of ​​the face image at different times over a period of time, and the absolute average difference between the brightness values ​​of the edge pixels and the neighboring edge pixels of each area, and calibrated as , and , Indicates that within a period of time Moment face image The average brightness value of all pixels in the area Indicates that within a period of time Moment face image The variance of the brightness values ​​of all pixels in the region is Indicates that within a period of time Moment The absolute average difference between the brightness values ​​of the edge pixels in the region and the edge pixels in the neighborhood, , , and All are positive integers; Calculate the brightness balance deviation index of each area. The specific calculation formula is as follows: In the formula, For the Brightness balance deviation index of each area; The texture structure characteristic information in the artifact quality assessment information of each area of ​​the preprocessed face image is extracted, including the average change rate, amplitude and directional distribution variance of the pixel gradient in each area of ​​the face image at different times over a period of time, and calibrated as , and , Indicates that within a period of time Moment face image The average rate of change of pixel gradients in a region is Indicates that within a period of time Moment face image The magnitude of the pixel gradient in the region, Indicates that within a period of time Moment face image The directional distribution variance of pixel gradients in a region; Calculate the detail texture destruction index of each area. The specific calculation formula is as follows: In the formula, For the The detail texture destruction index of the area; An artifact impact assessment model is constructed for the brightness balance deviation index and detail texture destruction index of each generated area, and the artifact impact coefficient of each area is generated. After the generation, the model is analyzed to evaluate the impact of the artifact on the quality of each area. According to the evaluation results, each area is divided into high-impact area, medium-impact area and low-impact area.

3. The facial image quality assessment method according to claim 2, characterized in that: Brightness balance deviation index for each generated area and detail texture destruction index Construct an artifact impact assessment model and generate the artifact impact coefficient of each area through weighted summation , and the artifact influence coefficients of each region generated The threshold interval of the pre-set artifact influence coefficient Compare and evaluate the impact of artifacts on the quality of each area based on the comparison results, and divide each area into high-impact area, medium-impact area and low-impact area based on the evaluation results. The specific comparison analysis and division are as follows: like , if the artifact has a low impact on the quality of the area, then the area is classified as a low-impact area; like , if the artifact has a moderate impact on the quality of the area, then the area is classified as a medium-impact area; like If the artifact has a high degree of impact on the quality of the area, the area is divided into a high-impact area.

4. The facial image quality assessment method according to claim 3, characterized in that: According to the division results of each area of ​​the face image, a dynamic intervention mechanism is constructed, specifically: according to the division results of the high-impact area, the medium-impact area and the low-impact area, different intervention measure parameters are set respectively to form a dynamic intervention mechanism; the dynamic intervention mechanism is based on the artifact influence coefficient of each area and automatically determines the adjustment method and amplitude of brightness adjustment, texture enhancement and edge smoothing through pre-set rules; Corresponding intervention measures are taken for high-impact areas, medium-impact areas and low-impact areas respectively. Specifically, in high-impact areas, the artifact suppression enhancement parameters in the dynamic intervention mechanism are used to significantly adjust the brightness distribution, strengthen texture enhancement and optimize edge smoothing; in medium-impact areas, the artifact moderate adjustment parameters in the dynamic intervention mechanism are used to moderately optimize the brightness distribution and texture characteristics; in low-impact areas, the artifact stabilization parameters in the dynamic intervention mechanism are maintained without adjustment, and only the regional status is monitored to maintain stable quality.

5. The method for assessing facial image quality according to claim 4, characterized in that: In the process of dynamic intervention mechanism intervening in various regions, the intervention feedback information of each region is obtained in real time, and analyzed after acquisition to evaluate whether the intervention effect of the dynamic intervention mechanism in each region can achieve the expected effect, and optimize the dynamic intervention mechanism according to the evaluation results, which specifically includes the following steps: In the process of dynamic intervention mechanism intervening in each area, the intervention feedback information of each area is obtained in real time and preprocessed after acquisition; Extracting brightness adjustment feedback information and texture optimization feedback information from the preprocessed intervention feedback information of each region, and analyzing them after acquisition to generate a brightness adjustment deviation index and a texture restoration consistency index for each region respectively; The logic for obtaining the brightness adjustment deviation index and texture restoration consistency index of each area is as follows: The brightness adjustment feedback information in the intervention feedback information of each area after preprocessing is extracted, specifically including the average value and variance of the pixel brightness values ​​of each area at different times within a period of time after the intervention, and the preset average value of the pixel brightness values ​​of each area after the intervention, and they are calibrated as , and , Indicates that within a period of time after the intervention Moment The average value of the pixel brightness in the region Indicates that within a period of time after the intervention Moment The variance of the pixel brightness values ​​in the region, The pre-specified intervention The average value of the pixel brightness in the region , , and All are positive integers; Calculate the brightness adjustment deviation index of each area. The specific calculation formula is as follows: In the formula, For the Brightness adjustment deviation index for each region; The texture optimization feedback information in the intervention feedback information of each region after preprocessing is extracted, including the average amplitude of the pixel gradient of each region at different times within a period of time after the intervention, the directional distribution variance, and the pre-set average amplitude of the pixel gradient of each region after the intervention, and they are calibrated as , and , Indicates that within a period of time after the intervention Moment The average magnitude of the pixel gradient in the region, Indicates that within a period of time after the intervention Moment The directional distribution variance of the pixel gradient in the region, The pre-specified intervention The average amplitude of the pixel gradient in the region; Calculate the texture restoration consistency index of each area. The specific calculation formula is as follows: In the formula, For the Texture recovery consistency index of the region; An intervention effect evaluation model is constructed for the brightness adjustment deviation index and texture restoration consistency index generated for each area, and the intervention coefficient of each area is generated. The generated intervention coefficient of each area is compared with the pre-set intervention coefficient threshold of each area. According to the comparison results, it is evaluated whether the intervention effect of the dynamic intervention mechanism in each area can meet the expectations, and the dynamic intervention mechanism is optimized according to the evaluation results.

6. The method for assessing facial image quality according to claim 5, characterized in that: Adjust the brightness of each area generated by the deviation index and texture recovery consistency index Construct an intervention effect evaluation model and generate intervention coefficients for each region through weighted summation , and the intervention coefficients of each region generated The intervention coefficient thresholds for each area are set in advance Compare and evaluate whether the intervention effect of the dynamic intervention mechanism in each region can meet expectations based on the comparison results, and optimize the dynamic intervention mechanism based on the evaluation results. The specific comparison and analysis are as follows: like , the intervention effect of the dynamic intervention mechanism in this area can achieve the expected effect, and there is no need to optimize the dynamic intervention mechanism; like The intervention effect of the dynamic intervention mechanism in this area cannot meet the expectations, and the dynamic intervention mechanism needs to be optimized, including: increasing the brightness adjustment intensity of the area and reducing the brightness distribution deviation; increasing the parameter weight of texture enhancement, strengthening the contrast of details and high-frequency texture characteristics; combining the real-time intervention feedback data of the area, dynamically adjusting the intervention parameters and rules, and optimizing the execution process of the intervention mechanism.

7. A facial image quality assessment system, used to implement the facial image quality assessment method according to any one of claims 1 to 6, characterized in that: It includes artifact effect detection module, area division and calibration module, artifact quality assessment module, dynamic intervention execution module, intervention effect optimization module and comprehensive monitoring improvement module; The artifact effect detection module performs artifact effect detection on the face image captured by the camera in real time under a dynamic light source environment. It determines whether there is an artifact effect by analyzing the uniformity of the image's brightness distribution, abnormal gradient direction, and changes in texture features. The region division and calibration module evenly divides the face image captured by the camera in real time into several regions through the facial key point detection method when it is determined that there is an artifact effect; The artifact quality assessment module obtains the artifact quality assessment information of each area of ​​the face image in real time, analyzes it after acquisition, evaluates the impact of the artifact on the quality of each area, and divides each area into high-impact area, medium-impact area and low-impact area according to the assessment results; The dynamic intervention execution module builds a dynamic intervention mechanism based on the division results of each area of ​​the face image, and takes corresponding intervention measures for the high-impact area, medium-impact area and low-impact area respectively; The intervention effect optimization module obtains the intervention feedback information of each area in real time during the process of the dynamic intervention mechanism intervening in each area, analyzes it after obtaining, evaluates whether the intervention effect of the dynamic intervention mechanism in each area can meet the expectations, and optimizes the dynamic intervention mechanism according to the evaluation results; The comprehensive monitoring and improvement module conducts real-time monitoring and comprehensive analysis of the artifact quality assessment information, intervention feedback information, intervention records and image quality restoration effects obtained during the artifact adjustment and evaluation process, continuously improves the artifact suppression method and dynamic intervention strategy, and optimizes the facial image quality assessment process.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the facial image quality assessment method according to any one of claims 1 to 6.

Images