A video quality detection method and system based on virtual production
By constructing a virtual production teaching environment and simulation analysis, the problem of subjectivity of manual review in virtual production video quality control is solved, and high-accurate video quality detection and evaluation are achieved, and the quality of the work is improved.
Patent Information
- Application Number
- CN202410511033.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-26
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-04-26
AI Technical Summary
In the quality control of virtual production videos, there are subjective judgment problems in manual review, which leads to inaccurate and reliable evaluation, lack of quantitative evaluation, and affects the improvement of the quality of the work.
By constructing a virtual production teaching environment, receiving production data input by users, generating virtual films, and performing simulation analysis, splitting them into multiple clips, comparing each clip with the target production data to generate simulation results, and generating simulation results Analytical reports provide quantitative matching degree and adjustment parameters.
It improves the accuracy and reliability of video quality inspection, provides objective quality assessment, helps to discover film differences and guides adjustments, and improves the overall quality of the work.
Smart Images

Figure CN118400585B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of virtual production teaching, and in particular to a video quality detection method and system based on virtual production. Background Art
[0002] Virtual Production is a process of creating and simulating the production of video content such as movies and TV series using computer graphics and virtual reality technology. It involves building and shooting scenes in a virtual environment, using virtual cameras, lighting systems and other virtual elements to create visual effects. In virtual production, traditional physical shooting is replaced by virtual scenes and characters, and the effects of the entire film can be designed and previewed through computer software without actually building scenes and actors.
[0003] In the quality control process of existing technologies, producers or quality control teams are required to watch the virtual production videos produced by users and evaluate their key parameters such as color, brightness, and contrast. Producers compare user works with known high-quality standard works to find differences. However, manual review and subjective judgment of quality control personnel may lead to inaccurate quality assessment results. Different personnel may have inconsistent quality judgment standards, which affects the reliability of the assessment results. In addition, there is a lack of quantitative assessment of the quality of the work through the personnel's quality judgment standards, and it is impossible to accurately point out the differences in each parameter. Users cannot adjust production data based on specific quantitative data, which affects the improvement of work quality.
[0004] Therefore, the prior art has defects and needs to be improved. Summary of the invention
[0005] In order to solve one or several problems in the prior art, the main purpose of the present application is to provide a video quality detection method and system based on virtual production.
[0006] In order to achieve the above-mentioned invention object, the present application proposes a video quality detection method based on virtual production, the method comprising:
[0007] Construct a virtual production teaching environment;
[0008] When receiving production data input by the user, inputting the production data into the virtual production teaching environment, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters;
[0009] When receiving the virtual filming video output by the virtual filming teaching environment, performing simulation analysis on the virtual filming video;
[0010] Splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data;
[0011] According to the results of the effect comparison, the matching degree between each clip and the target production data is output;
[0012] Based on the matching degree between the production data of each segment and the target production data, a simulation analysis report of the virtual filming is generated.
[0013] Furthermore, the comparison of the effect of the production data of each segment with the target production data includes analyzing the scene parameters of the production data of each segment, and the method includes:
[0014] Extract scene parameters of each clip production data;
[0015] Based on the scene parameters and the target production data, extracting the first color correction parameters of the scene parameters and extracting the second color correction parameters of the target production data;
[0016] Calculating a color difference of each pixel between the first color correction parameter and the second color correction parameter to obtain a color difference value, wherein the first color correction parameter and the second color correction parameter are used for the same color space of the production data and the target production data;
[0017] A degree of color matching of the scene parameters is determined based on the color difference value.
[0018] Furthermore, the effect comparison between the production data of each segment and the target production data includes analyzing the scene parameters of the production data of each segment, and the method further includes:
[0019] Extract scene parameters of each clip production data;
[0020] Inputting the scene parameters into a dynamic range analysis tool, performing brightness level analysis on the scene parameters using the dynamic range analysis tool, and outputting a distribution diagram of different brightness levels;
[0021] According to the brightness level distribution map, a contrast threshold is set for each brightness level in the distribution map. When the contrast of the brightness interval of the corresponding level is greater than the set contrast threshold, it is determined that details of the corresponding brightness interval are lost, and the corresponding brightness interval is marked as a focus area;
[0022] Performing tone mapping analysis on the portion marked as the area of interest, and analyzing the degree of detail preservation of the area of interest after tone mapping;
[0023] Based on the detail preservation degree, an effect evaluation of the scene parameters of each segment is completed.
[0024] Furthermore, the comparison of the effect of the production data of each segment with the target production data includes analyzing the crew parameters of the production data of each segment, including:
[0025] Extract crew parameters of each segment production data;
[0026] Extracting motion parameters of the crew parameters and motion parameters of target production data through a motion capture model;
[0027] Comparing and analyzing the action parameters of the crew parameters with the action parameters of the target production data to determine the matching degree of the action parameters;
[0028] Extracting facial feature variation parameters of crew members and facial feature variation parameters of target production data through a facial feature capture model;
[0029] Comparing and analyzing the facial feature change parameters of the crew member parameters and the facial feature change parameters of the target production data to determine the matching degree of the facial feature change parameters;
[0030] Based on the matching degree analysis results of the action parameters and the facial feature change parameters, the effect evaluation of the crew parameters of each clip is completed.
[0031] Furthermore, the comparison of the effect of the production data of each segment with the target production data includes analyzing the equipment parameters of the production data of each segment, including:
[0032] Extracting device parameters of each clip production data, wherein the device parameters include coordinate parameters, aperture size and exposure time of the camera;
[0033] The device parameters of each segment are compared and analyzed with the device parameters in the target production data, and the matching degree of the device parameters is determined according to the analysis result.
[0034] Furthermore, after generating a simulation analysis report of a virtual film based on the matching degree between the production data of each segment and the target production data, the method further includes:
[0035] Based on the matching degree between the production data of each segment and the target production data, the matching degree is quantified to obtain a matching value of the production data of the corresponding segment;
[0036] When the matching value is less than the preset matching threshold, the reason why the matching value is less than the preset matching threshold is analyzed, and an adjustment parameter is generated, wherein the adjustment parameter is used to adjust the production data to an interval parameter required for the target production data.
[0037] Furthermore, splitting the virtual film into multiple segments includes:
[0038] Acquire the scene parameters and crew parameters, classify the style of the virtual film according to the scene data and crew parameters, and formulate a splitting strategy for the virtual film according to the style classification;
[0039] According to the splitting strategy, a scene transition point is set for each scene, and the virtual filming movie is split into multiple segments according to the scene transition point.
[0040] The embodiment of the present application also provides a video quality detection system based on virtual production, including:
[0041] Construction module, used to build a virtual production teaching environment;
[0042] An input module, for inputting the production data into the virtual production teaching environment when receiving the production data input by the user, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters;
[0043] An analysis module, configured to perform simulation analysis on a virtual filming video when receiving the virtual filming video output by the virtual filming teaching environment;
[0044] A comparison module, used for splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data;
[0045] An output module is used to output the matching degree between each clip and the target production data according to the result of effect comparison;
[0046] The generation module is used to generate a simulation analysis report of the virtual filming film based on the matching degree between the production data of each segment and the target production data.
[0047] The present application also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.
[0048] The present application also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.
[0049] The video quality detection method and system based on virtual production of the embodiment of the present application can simulate various parameters and scenes in the real production process by constructing a virtual production teaching environment, and provide a controllable environment for subsequent video quality detection. Constructing a virtual production teaching environment can help realize virtual processing and simulation of production data, and improve the accuracy and reliability of video quality detection. By performing simulation analysis on the virtual filming film, the realism and simulation of the film can be objectively evaluated, providing a basis for subsequent effect comparison. Simulation analysis can help find the difference between the virtual filming film and the target production data, guide the subsequent quality assessment work, and improve the accuracy of the assessment results. By splitting the virtual filming film into multiple segments, and comparing the effect of the production data and the target production data for each segment, the quality problems of the film can be analyzed more carefully. Splitting and comparing each segment can help accurately find the problems in the film, guide the subsequent adjustment and optimization work, and improve the quality of the overall work. By generating a simulation analysis report, the assessment results of the film quality can be summarized and displayed, providing an intuitive reference for the production staff. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 A schematic diagram of a flow chart of a video quality detection method based on virtual production according to an embodiment of the present application;
[0051] Figure 2 A schematic diagram of a flow chart of a video quality detection method based on virtual production according to an embodiment of the present application;
[0052] Figure 3 A schematic block diagram of the structure of a video quality detection system based on virtual production according to an embodiment of the present application;
[0053] Figure 4 A schematic block diagram of the structure of a computer device according to an embodiment of the present application.
[0054] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0055] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0056] Reference Figure 1 In an embodiment of the present application, a video quality detection method based on virtual production is provided, the method comprising:
[0057] S1. Construct a virtual film production teaching environment;
[0058] S2. When receiving production data input by the user, input the production data into the virtual production teaching environment, and output a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters;
[0059] S3, when receiving the virtual filming video output by the virtual filming teaching environment, performing simulation analysis on the virtual filming video;
[0060] S4, splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data;
[0061] S5. Output the matching degree between each segment and the target production data according to the result of the effect comparison;
[0062] S6. Generate a simulation analysis report of the virtual filming video based on the matching degree between the production data of each segment and the target production data.
[0063] As described in the above steps S1-S2, computer graphics technology can be used to create a virtual scene environment, integrating elements required for virtual production such as virtual cameras, lighting systems, virtual actors, etc., providing a customizable virtual production environment, allowing users to perform pre-production and testing without actually building scenes and actors, reducing production costs and improving production efficiency. Users input production data such as crew members, equipment, and scenes through the interface, and the virtual production teaching environment applies these data to the virtual scene to generate a virtual film, realizing the rapid import and application of production data, improving production efficiency, and users can preview the virtual shooting effect based on the input data, discover and solve problems in advance. Through the virtual production teaching environment, a teaching platform for film and television production can be provided to users, and users can learn the production of film and television videos by changing production data.
[0064] As described in step S3 above, characteristic parameters of the virtual film, such as color, brightness, contrast, etc., can be extracted, and the virtual film can be analyzed for simulation using a preset evaluation model. This achieves a quantitative evaluation of the quality of the virtual film, improving the objectivity and accuracy of the evaluation. This helps users understand the differences between the virtual film and the real production, providing a basis for adjustment.
[0065] As described in step S4 above, the virtual film is split into multiple clips according to the scene transition, lens switching, action sequence and other standards, and each clip contains complete scene and production data. A refined analysis of the virtual film is achieved, and the quality of each clip can be evaluated and adjusted. The quality control level of the virtual film is improved. In particular, for certain styles, it can be split according to the scene transition points; for other styles, it can be split according to the duration of the scene. For example, for action scenes, it can be split according to the action sequence; for dialogue scenes, it can be split according to the dialogue paragraphs. Consider the special elements that may appear in the scene, such as transition effects, changes in lens language, etc., and adjust the splitting strategy accordingly.
[0066] As described in the above steps S5-S6, the production data of each clip is extracted, including crew parameters, equipment parameters and scene parameters. These data are compared with the target production data to evaluate their matching degree. A refined evaluation of the production data is achieved, and adjustments can be made for each parameter. The quality of virtual production works is improved. The evaluation results of each clip are summarized to generate a simulation analysis report of the virtual filming film. The report contains the matching degree and quality evaluation results of each clip. A detailed virtual production quality evaluation report is provided to the user, which helps the user understand the problems in the production process, helps the user make adjustments according to the suggestions in the report, and improves the quality of the virtual production works.
[0067] Specifically, by constructing a virtual production teaching environment, various parameters and scenes in the real production process can be simulated, providing a controllable environment for subsequent video quality detection. Constructing a virtual production teaching environment can help realize virtual processing and simulation of production data, and improve the accuracy and reliability of video quality detection. By analyzing the simulation degree of virtual filming, the realism and simulation degree of the film can be objectively evaluated, providing a basis for subsequent effect comparison. Simulation degree analysis can help find the difference between the virtual filming film and the target production data, guide the subsequent quality assessment work, and improve the accuracy of the assessment results. By splitting the virtual filming film into multiple segments and comparing the effect of the production data of each segment with the target production data, the quality problems of the film can be analyzed more carefully. Splitting and comparing each segment can help accurately find the problems in the film, guide the subsequent adjustment and optimization work, and improve the quality of the overall work. By generating a simulation degree analysis report, the evaluation results of the film quality can be summarized and displayed, providing an intuitive reference for producers.
[0068] In one feasible embodiment, a virtual scene is created, including buildings, streets, pedestrians and vehicles. A virtual camera and lighting system are set up to simulate a real production environment. Students input production data such as crew members, equipment, and scenes. The virtual production teaching environment generates a virtual film based on the input data. The virtual production teaching environment automatically performs simulation analysis on the generated virtual film. Evaluate key parameters such as color, brightness, and contrast of the film. According to scene transitions and lens switching, the virtual film is split into multiple clips. Each clip contains a complete scene and production data. The virtual production teaching environment automatically extracts the production data of each clip. These data are compared with the preset target production data to evaluate the degree of match. The virtual production teaching environment summarizes the evaluation results of each clip. Generate a simulation analysis report, including the matching degree and quality evaluation results of each clip. Students adjust the production data according to the suggestions in the report. Regenerate the virtual film and perform simulation analysis. When the production data of each clip is compared with the target production data, the specific comparison parameters can be: Delta E: a measure of color difference, ranging from 0 (perfect match) to infinity (complete mismatch). Color saturation: the vividness of the color, which can be quantified by a color saturation index (such as HSV or HSL). Brightness: the average brightness level of the image, which can be calculated by averaging the brightness value of each pixel. Contrast: a measure of brightness difference, which can be obtained by calculating the difference between the maximum and minimum brightness values in the image. Camera parameters: Focal length: the focal length of the camera lens, which affects the viewing angle and depth of field. Aperture: the aperture size of the camera lens, which affects the depth of field and the brightness of the image. Exposure time: the length of time the camera captures the image, which affects the brightness and motion blur of the image. Scene layout and design: Composition: the layout of objects in the image, including balance, alignment and focus. Scene depth: the sense of depth and hierarchy in the scene, which can be evaluated by analyzing the occlusion relationship between objects. Action synchronization: the degree of synchronization between the actions of the characters in the animation and the audio or plot. Expression and body language: The naturalness and accuracy of the virtual actor's expression and body language. Based on the above analysis, the quality of virtual production films is automatically analyzed and evaluated, avoiding manual review and subjective judgment of quality control personnel, and improving the accuracy and reliability of the evaluation results. Users can make targeted adjustments based on the adjustment suggestions and optimization plans in the simulation analysis report to improve the quality of virtual production works.
[0069] Reference Figure 2 In one embodiment, the comparison of the effect of the production data of each segment with the target production data includes analyzing the scene parameters of the production data of each segment, and the method includes:
[0070] S41, extracting scene parameters of each segment production data;
[0071] S42, based on the scene parameters and the target production data, extracting the first color correction parameters of the scene parameters and extracting the second color correction parameters of the target production data;
[0072] S43, calculating the color difference of each pixel between the first color correction parameter and the second color correction parameter to obtain a color difference value, wherein the first color correction parameter and the second color correction parameter are used for the same color space of the production data and the target production data;
[0073] S44: Determine a color matching degree of the scene parameter based on the color difference value.
[0074] As described in the above steps, computer vision and image processing techniques can be used to extract scene parameters, such as ambient lighting, color, texture, etc., from each segment of the virtual production film. This provides the necessary basic data for subsequent color matching analysis. Ensure that the scene parameters of each segment can be accurately extracted and analyzed. Analyze the scene parameters of each segment and determine the first color correction parameters corresponding to the scene parameters in the target production data. Ensure that the scene parameters of the virtual production film are consistent with the scene parameters in the target production data in color space. Provide key correction parameters for color matching analysis. Extract second color correction parameters corresponding to the first color correction parameters from the target production data. Ensure that the scene parameters of the virtual production film are consistent with the scene parameters in the target production data in color space. Provide key correction parameters for color matching analysis. Use color difference formulas (such as CIE76, CIE94, CIEDE2000, etc.) to calculate the color difference value of each pixel. Quantify the color difference between the scene parameters of the virtual production film and the target production data. Provide specific quantitative indicators for color matching analysis. Determine the color matching degree of the scene parameters based on the calculated color difference values. It provides users with a quantitative color matching index to help users understand the color difference between the virtual production film scene parameters and the target production data.
[0075] In one embodiment, the comparing the effect of the production data of each segment with the target production data includes analyzing the scene parameters of the production data of each segment, and the method further includes:
[0076] Extract scene parameters of each clip production data;
[0077] Inputting the scene parameters into a dynamic range analysis tool, performing brightness level analysis on the scene parameters using the dynamic range analysis tool, and outputting a distribution diagram of different brightness levels;
[0078] According to the brightness level distribution map, a contrast threshold is set for each brightness level in the distribution map. When the contrast of the brightness interval of the corresponding level is greater than the set contrast threshold, it is determined that details of the corresponding brightness interval are lost, and the corresponding brightness interval is marked as a focus area;
[0079] Performing tone mapping analysis on the portion marked as the area of interest, and analyzing the degree of detail preservation of the area of interest after tone mapping;
[0080] Based on the detail preservation degree, an effect evaluation of the scene parameters of each segment is completed.
[0081] As described above, scene parameters such as brightness, color, etc. are extracted from each clip of the virtual production film. This provides the necessary basic data for the subsequent brightness level analysis and tone mapping analysis. Ensure that the scene parameters of each clip can be accurately extracted and analyzed. Use professional dynamic range analysis tools, such as dynamic range image analysis software, to analyze the brightness level of scene parameters. Provide professional brightness level analysis tools to more accurately evaluate the distribution of brightness levels. Help users understand the brightness level distribution of virtual production film scene parameters. Analyze the brightness level distribution of scene parameters through dynamic range analysis tools. Quantify the brightness level distribution of virtual production film scene parameters. Help users understand the impact of brightness level distribution on detail retention. According to the brightness level distribution diagram, set the contrast threshold for each level. Set a contrast threshold for each brightness interval in the brightness level distribution to evaluate the degree of detail retention. Help users understand which brightness intervals may have lost details. When the contrast is greater than the set contrast threshold, it is determined that the details of the brightness interval are lost. Help users identify possible detail loss problems in the brightness level distribution. Mark the area of interest for further analysis. Perform tone mapping analysis on the parts marked as areas of interest to evaluate the degree of detail retention after tone mapping. Provides professional tone mapping analysis tools to more accurately evaluate the degree of detail retention after tone mapping. Helps users understand the impact of tone mapping on detail retention. Based on the results of tone mapping analysis, completes the effect evaluation of scene parameters for each clip. Quantifies the effect of virtual production film scene parameters to help users understand the pros and cons of scene parameters. Provides users with targeted optimization suggestions to improve the effect of scene parameters.
[0082] It is worth mentioning that if you only look at the parameters without considering the impact of each parameter on the details, it may lead to the following problems:
[0083] Suppose there is a virtual production teaching scenario where a student is making a virtual film of an outdoor scene. The student inputs the following production data:
[0084] Scene parameters: Outdoor scene, including grass, trees and sky.
[0085] Camera parameters: focal length 50mm, aperture f / 4, exposure time 1 / 125 second.
[0086] Lighting system: natural light, moderate light intensity.
[0087] Students used these parameters for virtual production and generated a virtual film of an outdoor scene. In traditional quality control methods, one might only focus on the camera parameters, exposure time, and lighting system parameters, while ignoring their impact on the details of the scene. For example, if the camera focal length is set to 50mm and the aperture is set to f / 4, this may result in the loss of details of the trees and grass in the scene, as these details may become blurred due to insufficient depth of field. However, if one simply looks at these parameters without considering their impact on the details of the scene, one might draw the following conclusions:
[0088] The camera focal length and aperture settings are reasonable, and the exposure time is moderate.
[0089] The lighting system provides appropriate natural lighting.
[0090] The scene parameters are set reasonably, and a virtual movie of an outdoor scene can be generated.
[0091] From the above analysis, the impact of camera parameters on scene details is ignored, resulting in students not being aware of the impact of camera settings on details and thus unable to make corresponding adjustments. This may cause the loss of tree and grass details in the virtual film produced by students, affecting the overall quality of the film. By considering the impact of each parameter on scene details, students can evaluate production data more comprehensively and make targeted adjustments. This helps improve the quality of virtual production works and helps students master virtual production techniques.
[0092] In one embodiment, the comparison of the effect of the production data of each segment with the target production data includes analyzing the crew parameters of the production data of each segment, including:
[0093] Extract crew parameters of each segment production data;
[0094] Extracting motion parameters of the crew parameters and motion parameters of target production data through a motion capture model;
[0095] Comparing and analyzing the action parameters of the crew parameters with the action parameters of the target production data to determine the matching degree of the action parameters;
[0096] Extracting facial feature variation parameters of crew members and facial feature variation parameters of target production data through a facial feature capture model;
[0097] Comparing and analyzing the facial feature change parameters of the crew member parameters and the facial feature change parameters of the target production data to determine the matching degree of the facial feature change parameters;
[0098] Based on the matching degree analysis results of the action parameters and the facial feature change parameters, the effect evaluation of the crew parameters of each clip is completed.
[0099] As described above, crew parameters, including motion and facial features, are extracted from each clip of the virtual production film. This provides the necessary basic data for subsequent motion and facial feature analysis. This ensures that crew parameters for each clip can be accurately extracted and analyzed. Using motion capture technology, such as optical motion capture or inertial motion capture, the actor's motion parameters are extracted from the virtual production film. This provides accurate actor motion data, which helps to evaluate the degree of motion matching. This helps users understand the authenticity and naturalness of the motion in the virtual production film. The extracted motion parameters are compared and analyzed with the motion parameters in the target production data. This quantifies the difference between the motion parameters in the virtual production film and the target production data. This helps users understand the degree of motion matching and provides a basis for motion adjustment. Using facial capture technology, such as 3D scanning or depth cameras, the actor's facial feature change parameters are extracted from the virtual production film. This provides accurate actor facial feature data, which helps to evaluate the degree of facial feature change matching. This helps users understand the naturalness and accuracy of facial features in the virtual production film. The extracted facial feature change parameters are compared and analyzed with the facial feature change parameters in the target production data. Quantified the difference between facial feature changes in virtual production videos and target production data. Helped users understand the matching degree of facial feature changes and provided a basis for facial adjustments. Completed the effect evaluation of crew parameters for each clip based on the matching degree analysis results of movements and facial feature changes. Provided an overall matching degree evaluation of crew parameters to help users understand the performance of actors in virtual production videos. Provided users with targeted adjustment suggestions to improve the quality and naturalness of crew members.
[0100] In a feasible embodiment, the action matching degree calculation is a complex process, which requires calculating the difference between the action parameters of the actors in the virtual filming film and the action parameters of the corresponding actors in the target production data, so as to obtain the action matching degree. Specifically: Extract action parameters: Extract key action parameters from the virtual action data set (A), including action trajectory, speed and acceleration. Extract corresponding key action parameters from the target action data set (B).
[0101] The formula for the motion trajectory difference is: d_trajectory = sqrt[(x_A-x_B)^2+(y_A-y_B)^2+(z_A-z_B)^2], where (x_A, y_A, z_A) and (x_B, y_B, z_B) are the coordinates of a motion trajectory point in the virtual motion dataset and the target motion dataset respectively, d_trajectoryd is the distance of the motion trajectory difference, and the result is (A, B) is the distance coordinate of the motion trajectory difference.
[0102] The formula of speed difference is: v_dif=|v_A-v_B|, where v_A and v_B are the speeds of one of the key frames in the virtual action dataset and the target action dataset, respectively.
[0103] The formula of acceleration difference is: a_dif=|a_A-a_B|, where a_A and a_B are the acceleration of one of the key frames in the virtual action dataset and the target action dataset, respectively.
[0104] Comprehensive matching degree calculation: The difference values of position, velocity and acceleration are weighted averaged to obtain the comprehensive matching degree. The weight is determined according to the importance of the action parameters, and w1, w2, and w3 are calculated, where w1, w2, and w3 are the weights of the action trajectory, velocity, and acceleration, respectively.
[0105] Motion matching formula: M = (w1*d_trajectory+w2*v_dif+w3*a_dif) / (w1+w2+w3), where w1, w2, w3 are the weights of motion trajectory, velocity and acceleration, and d_trajectory, v_dif, a_dif are the difference values of the above corresponding parameters.
[0106] From the above analysis, we can get the difference between the action parameters of the actors in the virtual film and the action parameters of the corresponding actors in the target production data, and thus get the action matching degree. Through quantitative calculation, we can objectively evaluate the matching degree between the virtual action and the target action, avoiding the subjective judgment in manual review. Users can accurately adjust the virtual action parameters according to the calculation results to improve the matching degree with the target action. This helps to improve the quality of virtual production works. Through calculation, the difference between action parameters, including position, speed, acceleration, etc., can be accurately evaluated. Small differences in these parameters may have a significant impact on the realism of the action, and it is difficult for the naked eye to capture these details. The calculation method reduces the interference of subjective judgment because it is based on quantitative data. Different observers may have different views on the authenticity of the action, and the calculation method provides a unified evaluation standard.
[0107] In one embodiment, the comparison of the effect of the production data of each segment with the target production data includes analyzing the device parameters of the production data of each segment, including:
[0108] Extracting device parameters of each clip production data, wherein the device parameters include coordinate parameters, aperture size and exposure time of the camera;
[0109] The device parameters of each segment are compared and analyzed with the device parameters in the target production data, and the matching degree of the device parameters is determined according to the analysis result.
[0110] As described above, the device parameters are extracted from each clip of the virtual production film, including the coordinate parameters, aperture size, and exposure time of the camera. This provides the necessary basic data for the subsequent comparative analysis of the device parameters. This ensures that the device parameters of each clip can be accurately extracted and analyzed. This compares and analyzes the extracted device parameters with the device parameters in the target production data. This quantifies the difference between the device parameters in the virtual production film and the target production data. This helps users understand whether the settings of the device parameters in the virtual production film meet expectations. Based on the results of the comparative analysis, the matching degree of the device parameters is determined. This provides an evaluation of the matching degree of the device parameters to help users understand the pros and cons of the device parameters in the virtual production film. This provides users with targeted optimization suggestions to improve the matching degree of the device parameters.
[0111] In one embodiment, after generating a simulation analysis report of a virtual film based on the matching degree between the production data of each segment and the target production data, the method further includes:
[0112] Based on the matching degree between the production data of each segment and the target production data, the matching degree is quantified to obtain a matching value of the production data of the corresponding segment;
[0113] When the matching value is less than the preset matching threshold, the reason why the matching value is less than the preset matching threshold is analyzed, and an adjustment parameter is generated, wherein the adjustment parameter is used to adjust the production data to an interval parameter required for the target production data.
[0114] As described above, the production data of each clip is extracted, including crew members, equipment and scene parameters. The difference between these parameters and the target production data is calculated, and the degree of matching is evaluated using quantitative indicators (such as color difference, position error, action matching, etc.). The difference value of each parameter is converted into a percentage or score to obtain a matching value. A quantitative evaluation of the degree of matching of the production data is provided, making the evaluation results more objective and quantifiable. Help users understand the degree of difference between the production data and the target production data. Analyze each clip with a matching value less than a preset threshold to determine the cause of the low matching. This may include analysis of color, brightness, contrast, camera parameters, actor movements, etc. Help users understand the specific reasons for the low matching of the production data and provide a basis for adjustment. Users can make targeted adjustments based on the analysis results. Generate corresponding adjustment parameters based on the reasons for the low matching. These parameters may include color temperature values for color correction, percentages for brightness adjustment, specific coordinates of the camera position, etc. Provide users with specific adjustment guidance to help them improve the matching degree of production data with target production data. Users can make adjustments based on the adjustment parameters to improve the quality of virtual production works.
[0115] In another embodiment, if the reason why the match value is less than the preset match threshold is that the degree of detail preservation is low, the method steps include: analyzing the degree of detail preservation of each segment to determine which details are lost or blurred in the virtual production. Using image processing techniques, such as edge detection and feature extraction, to identify and quantify details. Helping the user understand the degree of detail preservation. Analyzing the reasons for detail loss, such as camera settings, lighting conditions, scene layout, etc. Helping the user understand the specific reasons for detail loss. Generating adjustment parameters based on the reasons for detail loss, which may include camera focal length, aperture size, exposure time, light intensity, etc. Helping the user to improve the degree of detail preservation. The user can make adjustments based on the adjustment parameters to improve the quality of the virtual production work. The user adjusts the device settings in the virtual production environment based on the generated adjustment parameters. The user can make targeted adjustments to improve the degree of detail preservation. After adjusting the device settings, re-analyzing the degree of detail preservation. The user can provide feedback to indicate whether further adjustments are needed. Verifying the adjustment effect to ensure that the degree of detail preservation is as expected. Making further adjustments based on the feedback to improve the quality of the virtual production work. The user re-produces based on the adjusted device settings and re-analyzing the degree of detail preservation. Repeat the above steps until the level of detail preservation reaches a satisfactory level. Iteratively optimize the level of detail preservation to improve the quality of virtual production works.
[0116] In one embodiment, splitting the virtual film into multiple segments includes:
[0117] Acquire the scene parameters and crew parameters, classify the style of the virtual film according to the scene data and crew parameters, and formulate a splitting strategy for the virtual film according to the style classification;
[0118] According to the splitting strategy, a scene transition point is set for each scene, and the virtual filming movie is split into multiple segments according to the scene transition point.
[0119] As described above, scene data, such as scene layout, ambient lighting, color, etc., are extracted from each scene of the virtual production film. At the same time, crew parameters, such as the position, action, expression, etc. of the actors, are extracted. This provides the necessary basic data for the subsequent style classification and splitting strategy. This ensures that each scene and crew parameter can be accurately extracted and analyzed. Using machine learning or deep learning algorithms, the virtual film is style-classified based on the scene data and crew parameters. These algorithms can identify and classify different style features, such as color tone, composition, action style, etc. A style classification of the virtual film is provided to help users understand the style features of the film. This provides a basis for the splitting strategy to better organize and manage the virtual film. Based on the style classification results, a splitting strategy is formulated. For example, for action style, it may be necessary to split according to action sequence; for dialogue style, it may be necessary to split according to dialogue paragraphs. Splitting strategies for different style features are provided to help users better organize and manage virtual film. This helps users improve the quality of virtual production works. According to the splitting strategy, scene transition points are set for each scene. These transition points can be determined based on style features, action sequences, dialogue paragraphs, etc. It provides accurate splitting basis, which helps users better organize and manage virtual filming videos and improve the quality of virtual production works.
[0120] The video quality detection method based on virtual production of the present application can simulate various parameters and scenes in the real production process by constructing a virtual production teaching environment, and provide a controllable environment for subsequent video quality detection. Constructing a virtual production teaching environment can help realize virtual processing and simulation of production data, and improve the accuracy and reliability of video quality detection. By analyzing the degree of simulation of virtual filming, the realism and degree of simulation of the film can be objectively evaluated, providing a basis for subsequent effect comparison. Simulation analysis can help find the difference between the virtual filming film and the target production data, guide the subsequent quality assessment work, and improve the accuracy of the assessment results. By splitting the virtual filming film into multiple segments, and comparing the effect of the production data and the target production data for each segment, the quality problems of the film can be analyzed more carefully. Splitting and comparing each segment can help accurately find the problems in the film, guide the subsequent adjustment and optimization work, and improve the quality of the overall work. By generating a simulation analysis report, the evaluation results of the film quality can be summarized and displayed, providing an intuitive reference for the production staff.
[0121] Reference Figure 3 In an embodiment of the present application, a video quality detection system based on virtual production is also provided, comprising:
[0122] Construction module 1, used to build a virtual production teaching environment;
[0123] Input module 2, for inputting the production data into the virtual production teaching environment when receiving the production data input by the user, and outputting the virtual filming film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters;
[0124] An analysis module 3 is used for performing simulation analysis on the virtual filming film when receiving the virtual filming film output by the virtual filming teaching environment;
[0125] A comparison module 4 is used to split the virtual film into multiple segments and compare the production data of each segment with the target production data;
[0126] Output module 5, used for outputting the matching degree between each segment and the target production data according to the result of effect comparison;
[0127] The generating module 6 is used to generate a simulation analysis report of the virtual filming film based on the matching degree between the production data of each segment and the target production data.
[0128] As described above, it can be understood that the various components of the video quality detection system based on virtual production proposed in this application can realize the functions of any one of the video quality detection methods based on virtual production described above, and the specific structure will not be repeated.
[0129] Reference Figure 4 In the embodiment of the present application, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as monitoring data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a video quality detection method based on virtual production is implemented.
[0130] The processor executes the video quality detection method based on virtual production, including: constructing a virtual production teaching environment; when receiving production data input by a user, inputting the production data into the virtual production teaching environment, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters; when receiving the virtual film output by the virtual production teaching environment, performing a simulation analysis on the virtual film; splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data; according to the result of the effect comparison, outputting the matching degree of each segment with the target production data; and generating a simulation analysis report of the virtual film based on the matching degree of the production data of each segment with the target production data.
[0131] The above-mentioned video quality detection method based on virtual production can simulate various parameters and scenes in the real production process by constructing a virtual production teaching environment, and provide a controllable environment for subsequent video quality detection. Constructing a virtual production teaching environment can help realize virtual processing and simulation of production data, and improve the accuracy and reliability of video quality detection. By analyzing the simulation degree of virtual filming, the realism and simulation degree of the film can be objectively evaluated, providing a basis for subsequent effect comparison. Simulation degree analysis can help find the difference between the virtual filming film and the target production data, guide the subsequent quality evaluation work, and improve the accuracy of the evaluation results. By splitting the virtual filming film into multiple segments and comparing the effect of the production data of each segment with the target production data, the quality problems of the film can be analyzed more carefully. Splitting and comparing each segment can help accurately find the problems in the film, guide the subsequent adjustment and optimization work, and improve the quality of the overall work. By generating a simulation degree analysis report, the evaluation results of the film quality can be summarized and displayed, providing an intuitive reference for producers.
[0132] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, a video quality detection method based on virtual production is implemented, including the steps of: constructing a virtual production teaching environment; when production data input by a user is received, inputting the production data into the virtual production teaching environment, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters; when the virtual film output by the virtual production teaching environment is received, performing a simulation analysis on the virtual film; splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data; according to the result of the effect comparison, outputting the matching degree of each segment with the target production data; based on the matching degree of the production data of each segment with the target production data, generating a simulation analysis report of the virtual film.
[0133] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0134] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, device, article or method including the element.
[0135] The above description is only a preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A video quality detection method based on virtual production, characterized in that: The method comprises: Construct a virtual production teaching environment; When receiving production data input by the user, inputting the production data into the virtual production teaching environment, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters; When receiving the virtual filming video output by the virtual filming teaching environment, performing simulation analysis on the virtual filming video; Splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data; According to the results of the effect comparison, the matching degree between each clip and the target production data is output; Generate a simulation analysis report of the virtual filming based on the matching degree between the production data of each clip and the target production data; The effect comparison between the production data of each segment and the target production data includes analyzing the scene parameters of the production data of each segment, and the method includes: Extract scene parameters of each clip production data; Based on the scene parameters and the target production data, extracting the first color correction parameters of the scene parameters and extracting the second color correction parameters of the target production data; Calculating a color difference of each pixel between the first color correction parameter and the second color correction parameter to obtain a color difference value, wherein the first color correction parameter and the second color correction parameter are used for the same color space of the production data and the target production data; determining a degree of color matching of the scene parameters based on the color difference value; The effect comparison between the production data of each segment and the target production data includes analyzing the scene parameters of the production data of each segment. The method further includes: Extract scene parameters of each clip production data; Inputting the scene parameters into a dynamic range analysis tool, performing brightness level analysis on the scene parameters using the dynamic range analysis tool, and outputting a distribution diagram of different brightness levels; According to the brightness level distribution map, a contrast threshold is set for each brightness level in the distribution map. When the contrast of the brightness interval of the corresponding level is greater than the set contrast threshold, it is determined that details of the corresponding brightness interval are lost, and the corresponding brightness interval is marked as a focus area; Performing tone mapping analysis on the portion marked as the area of interest, and analyzing the degree of detail preservation of the area of interest after tone mapping; Based on the detail retention degree, completing effect evaluation of scene parameters of each segment; The effect comparison between the production data of each segment and the target production data includes analyzing the crew parameters of the production data of each segment, including: Extract crew parameters of each segment production data; Extracting motion parameters of the crew parameters and motion parameters of target production data through a motion capture model; Comparing and analyzing the action parameters of the crew parameters with the action parameters of the target production data to determine the matching degree of the action parameters; Extracting facial feature variation parameters of crew members and facial feature variation parameters of target production data through a facial feature capture model; Comparing and analyzing the facial feature change parameters of the crew member parameters and the facial feature change parameters of the target production data to determine the matching degree of the facial feature change parameters; Based on the matching degree analysis results of the action parameters and the facial feature change parameters, the effect evaluation of the crew parameters of each clip is completed.
2. The video quality detection method based on virtual production according to claim 1, characterized in that: After generating a simulation analysis report of a virtual film based on the matching degree between the production data of each segment and the target production data, the method further includes: Based on the matching degree between the production data of each segment and the target production data, the matching degree is quantified to obtain a matching value of the production data of the corresponding segment; When the matching value is less than the preset matching threshold, the reason why the matching value is less than the preset matching threshold is analyzed, and an adjustment parameter is generated, wherein the adjustment parameter is used to adjust the production data to an interval parameter required for the target production data.
3. The video quality detection method based on virtual production according to claim 1, characterized in that: The step of splitting the virtual film into multiple segments includes: Acquire the scene parameters and crew parameters, classify the style of the virtual film according to the scene data and crew parameters, and formulate a splitting strategy for the virtual film according to the style classification; According to the splitting strategy, a scene transition point is set for each scene, and the virtual filming movie is split into multiple segments according to the scene transition point.
4. The video quality detection method based on virtual production according to claim 1, characterized in that: The effect comparison between the production data of each segment and the target production data includes analyzing the equipment parameters of the production data of each segment, including: Extracting device parameters of each clip production data, wherein the device parameters include coordinate parameters, aperture size and exposure time of the camera; The device parameters of each segment are compared and analyzed with the device parameters in the target production data, and the matching degree of the device parameters is determined according to the analysis result.
5. A video quality detection system based on virtual production, used in the method according to any one of claims 1 to 4, characterized in that: include: Construction module, used to build a virtual production teaching environment; An input module, for inputting the production data into the virtual production teaching environment when receiving the production data input by the user, and outputting a virtual film through the virtual production teaching environment, wherein the production data includes crew parameters, equipment parameters and scene parameters; An analysis module, configured to perform simulation analysis on a virtual filming video when receiving the virtual filming video output by the virtual filming teaching environment; A comparison module, used for splitting the virtual film into multiple segments, and comparing the production data of each segment with the target production data; An output module is used to output the matching degree between each clip and the target production data according to the result of effect comparison; The generation module is used to generate a simulation analysis report of the virtual filming film based on the matching degree between the production data of each segment and the target production data.
6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.
