A method and system for intelligent perception and association of multimedia content in live sports events

Abstract

The present application provides a method and system for intelligent perception and association of multimedia content in live sports events. Among them, the original video data is obtained, and the color adjustment parameters are generated to complete the color space conversion in combination with the light reflection characteristics of different camera positions and the changes in the color temperature of the venue lighting. The priority of the coding block is dynamically adjusted according to the instantaneous speed and texture complexity of the moving object, and the coding quality parameters are adaptively compensated based on the brightness distribution of the bright area. Through two-way data interaction, the color component weight ratio is used to regulate the code stream distribution, and finally the processed picture is associated with the audio time series, motion trajectory and multi-camera space coordinates to form a live content transmission sequence. The technical solution provided by the present application realizes dynamic color optimization and improves coding efficiency, ensuring the simultaneous enhancement of live streaming media image quality and transmission stability under high-motion pictures.

Description

Technical Field

[0001] The present application relates to the field of intelligent perception and association of multimedia content, and in particular to a method and system for intelligent perception and association of multimedia content of live sports events. Background Art

[0002] Live broadcasts of sports events need to address issues such as lag in dynamic switching of multiple camera images and poor adaptability of content presentation across devices. These require intelligent perception and association methods for multimedia content that can: parse the semantics of group behavior in event footage (such as offensive and defensive formation transitions and team collaboration intentions) and predict the optimal perspective switching path; adaptively reconstruct data visualization levels based on terminal device display characteristics (such as mobile screen size and VR headset interaction logic) to achieve a cross-platform immersive viewing experience.

[0003] The current mainstream solution adopts a dynamic rendering architecture based on the collaboration of spatiotemporal knowledge graphs and edge computing: graph neural networks are used to model the spatiotemporal dependency between video stream frame sequences and IoT sensor data, and the inference engine deployed on edge nodes synchronously calculates the multimodal data fusion weights. The audience attention heat map is used as the reward function, and cloud-edge collaborative rendering resources are dynamically allocated to ultimately generate a device-adaptive enhanced viewing stream.

[0004] However, the construction of spatiotemporal knowledge graphs relies on a pre-set tactical rule base, making it difficult to capture temporary tactical innovations (such as impromptu teamwork in a football match). This leads to a discrepancy between dynamic perspective switching strategies and actual on-field intentions. Differences in heterogeneous data processing latency across edge computing nodes can easily lead to spatiotemporal misalignment in multimodal data overlays. Furthermore, reinforcement learning models are limited in capturing audience attention patterns through explicit interactions (such as pause / replay operations), making it difficult to quantify implicit cognitive load (such as the depth of tactical understanding). This leads to an imbalance in resource allocation for rendering complex VR scenes and exacerbates user nausea. Summary of the Invention

[0005] The present application provides a method and system for intelligently perceiving and associating multimedia content of live sports events, so as to solve the problems of unstable picture quality and poor correlation of multi-source data in the prior art.

[0006] In a first aspect, the present application provides a method for intelligently sensing and associating multimedia content of a live sports event, comprising:

[0007] Get the original video data;

[0008] Generate color adjustment parameters that match the color range distribution of the current motion picture based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data;

[0009] Adjusting the priority of coding block division based on the instantaneous motion speed and image texture complexity of moving objects in live sports events, and adaptively compensating coding quality parameters based on the brightness distribution of bright areas in the target video data;

[0010] Performing bidirectional data interaction according to the color adjustment parameter and the compensated encoding quality parameter, so that the color component weight ratio in the color adjustment parameter affects the code stream distribution ratio of the color component during the video encoding process;

[0011] Based on the target video data and the two-way data interaction results, a live sports event picture is constructed, and the live sports event picture is associated with the audio time series data, motion trajectory data and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence.

[0012] Optionally, performing bidirectional data interaction according to the color adjustment parameter and the compensated encoding quality parameter so that the color component weight ratio in the color adjustment parameter affects the code stream distribution ratio of the color component during the video encoding process includes:

[0013] Extracting weight ratio values ​​of red, green, and blue color components from the color adjustment parameter, and extracting compression rate limit values ​​corresponding to different coding block priority levels from the compensated coding quality parameter;

[0014] The weight ratio values ​​of the red, green, and blue color components are respectively superimposed with the compression rate limit values ​​of the corresponding coding block priority according to the preset weight ratio to generate the adaptation relationship parameters between the color component weight and the compression rate;

[0015] According to the numerical value of the adaptation relationship parameter, a positive correlation between the color component weight ratio value and the compression rate limit value and the bitstream distribution ratio is set to generate a bitstream distribution ratio mapping rule;

[0016] During the video encoding process, detecting a change in the weight ratio of the red, green, and blue color components in the color adjustment parameter, and adjusting the code stream allocation ratio of each color component according to the code stream allocation ratio mapping rule;

[0017] Comparing the adjusted bitrate allocation ratio with the brightness distribution data of the bright area in the compensated coding quality parameter frame by frame to generate a bitrate allocation deviation feedback signal;

[0018] According to the code stream allocation deviation feedback signal, the weight ratio of the red, green and blue color components in the color adjustment parameter is reversely corrected to form a two-way data interaction closed loop of the color component weight and the code stream allocation ratio, and generate a two-way data exchange result.

[0019] Optionally, based on the light reflection characteristics of different camera positions in a live sports event scene and the color temperature changes of the field lighting, color adjustment parameters matching the color range distribution of the current sports picture are generated, and the color space of the original video data is nonlinearly converted using the color adjustment parameters to obtain target video data, including:

[0020] Based on the light reflection characteristics of different camera positions in the live sports event scene, the light intensity distribution curve and reflection angle range of each camera position are extracted. At the same time, the color temperature values ​​of the venue lighting color temperature changes are collected to generate the current light reflection and color temperature parameter group for each camera position;

[0021] Extracting a distribution histogram of the red, green, and blue color components of the current motion picture from the original video data, and calculating the adaptation weights of the red, green, and blue color components by combining the reflection angle range and the color temperature value in the light reflection and color temperature parameter group;

[0022] Scaling the distribution histograms of the red, green, and blue color components based on the adaptation weights to generate color adjustment parameters that match the color range distribution of the current motion picture, the color adjustment parameters including nonlinear offset coefficients and gain coefficients of the red, green, and blue color components;

[0023] The red, green and blue color components of the original video data are subjected to piecewise nonlinear mapping through the nonlinear offset coefficient and gain coefficient in the color adjustment parameters, and the original color space is converted into the target color space to generate target video data.

[0024] Optionally, performing piecewise nonlinear mapping on the red, green, and blue color components of the original video data by using the nonlinear offset coefficient and the gain coefficient in the color adjustment parameters to convert the original color space into the target color space to generate the target video data, including:

[0025] Dividing the value range of each color component into a low value interval, a medium value interval, and a high value interval according to the distribution histogram value range of the red, green, and blue color components in the original video data, with the division boundaries aligned with the peak and valley values ​​of the histogram;

[0026] Binding the nonlinear offset coefficient and the gain coefficient to the low value interval, the middle value interval, and the high value interval of each color component, respectively, and setting a segmented mapping rule in which the low value interval is bound to the high gain coefficient and the positive offset coefficient, and the high value interval is bound to the low gain coefficient and the negative offset coefficient;

[0027] For the values ​​of the red, green, and blue color components of each pixel in the original video data, calling the bound segment mapping rule according to the value interval to which the value belongs, and mapping the target color component value;

[0028] The mapped target color component values ​​are superimposed according to the scaling rule of the distribution histogram to generate superimposed values, and the superimposed values ​​are used as the color value of each pixel in the target video data to generate the target video data.

[0029] Optionally, according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, the priority of the coding block division is adjusted, and the coding quality parameters are adaptively compensated in combination with the brightness distribution of the bright area in the target video data, including:

[0030] Extract the instantaneous speed of moving objects from live sports footage, calculate the displacement difference between adjacent video frames, and simultaneously extract the edge sharpness change of the image texture complexity to generate a joint evaluation parameter for motion speed and texture complexity.

[0031] Adjusting the priority of coding block division according to the product relationship between the displacement difference and the edge sharpness change in the joint evaluation parameter to generate a coding block priority distribution map for high motion areas and high texture areas;

[0032] Extracting the brightness distribution of bright areas from the target video data, calculating the area ratio of areas where the brightness value exceeds a preset highlight threshold, and generating a brightness distribution weight coefficient;

[0033] According to the priority levels of different areas in the coding block priority distribution map and in combination with the brightness distribution weight coefficient, weighted compensation is performed on the compression strength and detail retention ratio in the coding quality parameters to generate adaptively compensated coding quality parameters.

[0034] Optionally, based on the target video data and the two-way data interaction result, a live sports event picture is constructed, and the live sports event picture is associated with the audio time series data, motion trajectory data, and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence, including:

[0035] Extracting a video frame sequence after color space conversion from the target video data, and extracting a color component weight ratio, a bit stream allocation ratio parameter, and an encoding quality compensation parameter from a two-way data interaction result, and combining them with the video frame sequence to generate a live picture construction parameter group;

[0036] According to the color component weight ratio in the live picture construction parameter group, the numerical distribution of the red, green, and blue color components in the video frame is adjusted, and the bitstream resources of each frame are dynamically allocated in combination with the bitstream allocation ratio parameter to generate a live sports event picture including coding quality compensation;

[0037] Align each time node of the audio time series data in the live stream with the video frame timestamp of the live sports event screen frame by frame to generate synchronization timestamp parameters for the audio and video frames;

[0038] According to the position coordinates of the athletes in the motion trajectory data in the live stream, the corresponding view angles in the multi-camera position space coordinates are matched to obtain the spatial anchor point parameters of the motion trajectory and camera view angle in each video frame;

[0039] The sports event live broadcast images, synchronization timestamp parameters and spatial anchor point parameters are data-associated and packaged according to a preset transmission format to generate a live content transmission sequence.

[0040] Optionally, the position coordinates of the athletes in the motion trajectory data in the live stream are matched with the corresponding view angles in the multi-camera position space coordinates to obtain the spatial anchor point parameters of the motion trajectory and the camera view angle in each video frame, including:

[0041] Extracting the position coordinates of the athlete corresponding to the timestamp of each video frame from the motion trajectory data in the live stream, and generating a sequence of the athlete's position coordinates based on the position coordinates;

[0042] Extracting the spatial coordinates of each camera and the coverage of its shooting angle of view from the spatial coordinates of multiple camera positions;

[0043] Performing position matching on each coordinate point in the player position coordinate sequence with the coverage range, and if the coordinate point is within the coverage range, marking the camera as a matching camera, thereby obtaining a preliminary matching result;

[0044] Generating position mapping parameters of the relative pixel coordinates of the athlete's position in the camera image according to the relative position of each coordinate point in the preliminary matching result and the spatial coordinates of the matching camera;

[0045] The position mapping parameters are bound to the spatial coordinates of the matching camera to obtain spatial anchor point parameters.

[0046] In a second aspect, the present application provides a system for intelligently sensing and associating multimedia content of live sports events, including:

[0047] An acquisition module, used to obtain original video data;

[0048] A conversion module is configured to generate color adjustment parameters that match the color range distribution of the current sports image based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data;

[0049] An adjustment module is used to adjust the priority of coding block division according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, and to adaptively compensate the coding quality parameters based on the brightness distribution of the bright areas in the target video data;

[0050] An interaction module, configured to perform bidirectional data interaction based on the color adjustment parameters and the compensated encoding quality parameters, so that the color component weight ratio in the color adjustment parameters affects the code stream distribution ratio of the color components during video encoding;

[0051] A construction module is used to construct a live sports event screen based on the target video data and the two-way data interaction results, and to associate the live sports event screen with the audio time series data, motion trajectory data and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence.

[0052] In a third aspect, the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for intelligent perception and association of multimedia content of a live sports event as described in the first aspect above.

[0053] In a fourth aspect, the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements the method for intelligent perception and association of multimedia content of a live sports event as described in the first aspect.

[0054] By acquiring the original video data, the present application can establish the original information input channel of the live broadcast picture of the sports event; by generating color adjustment parameters that match the color range distribution of the current motion picture according to the light reflection characteristics of different camera positions in the live broadcast scene of the sports event and the changes in the color temperature of the venue lighting, and converting the color space of the original video data according to the color adjustment parameters to obtain the target video data, it can eliminate the color distortion caused by the differences in light sources of multiple cameras and enhance the visual layering of the moving objects; by adjusting the priority of the coding block division according to the instantaneous motion speed of the moving objects in the live broadcast picture of the sports event and the complexity of the picture texture, and combining the brightness distribution of the bright area in the target video data Adaptive compensation of coding quality parameters can improve the coding block boundary matching accuracy in dynamic scenes and reduce the quantization noise in highlight areas; by performing two-way data interaction based on the color adjustment parameters and the compensated coding quality parameters, the color component weight ratio affects the code stream allocation ratio of the color component during the video encoding process, which can achieve coordinated optimization of color gamut space conversion and coding compression algorithm; by constructing a live broadcast picture of a sports event based on the target video data and the results of two-way data interaction, and associating it with audio time series data, motion trajectory data and multi-camera position spatial coordinates to form a live content transmission sequence, it can achieve spatiotemporal synchronous mapping of multimodal data and enhance the immersive experience of multi-perspective live broadcast.

[0055] Furthermore, by extracting the weight ratio values ​​of the red, green and blue color components and superimposing and fusing them with the coding block compression rate limit values, a dynamic adaptive relationship between the color component weights and the compression rate is established, thereby realizing the joint optimization of color gamut conversion and coding compression; based on the positive correlation, the code stream allocation ratio mapping rules are generated, and the code stream allocation is dynamically adjusted in combination with the changes in the color component weights, effectively balancing color fidelity and code rate utilization; the color component weights are reversely corrected through the code stream allocation deviation feedback signal, forming a two-way data interaction closed loop, eliminating the spatial mismatch between chromaticity information and brightness distribution during the encoding process; finally, through frame-by-frame comparison and dynamic compensation mechanism, the color component code stream redundancy in high-dynamic images is reduced, the compression efficiency is improved while maintaining the integrity of the edge details of moving objects, so that the multi-camera live broadcast images can achieve adaptive coordination of color consistency and coding quality under complex lighting conditions, providing high-precision data support for multi-view low-latency live broadcast.

[0056] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0058] Figure 1 A flowchart of a method for intelligently sensing and associating multimedia content of a live sports event provided by the present application is shown;

[0059] Figure 2 A schematic diagram showing the structure of a system for intelligent perception and association of multimedia content of live sports events provided by the present application is shown;

[0060] Figure 3 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0061] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0062] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0063] Researchers have found that existing video processing technologies for live sports broadcasts lack synergy with dynamic lighting changes and motion picture coding efficiency, resulting in color distortion, rigid bitrate allocation, and delayed multi-source data fusion. Based on this, a dynamic optimization method for live sports broadcasts is provided. This method achieves dynamic adaptation of color reproduction, coding efficiency, and multi-data synchronization through bidirectional interaction between light reflection characteristics and coding parameters. The technical solution of this application is applicable to scenarios such as live sports broadcasts and multi-camera collaborative shooting, where both image quality and transmission stability must be taken into consideration.

[0064] The entire R&D process embodies a multi-dimensional collaborative optimization mechanism that links dynamic lighting adaptation with coding parameters, aiming to overcome the defects of existing solutions, such as color reproduction distortion, static bitrate allocation, and fragmentation of audio and video data. By analyzing light reflection characteristics and color temperature changes, the adaptability limitations of traditional color correction to dynamic scenes are overcome; the priority of coding blocks is dynamically adjusted based on motion speed and texture complexity to address the problem of picture quality fluctuations caused by fixed coding strategies; the interaction between color component weights and bitrate allocation is utilized to enhance the synergy between color fidelity and transmission efficiency; and accurate matching of audio and video trajectories is achieved through multi-source data association and spatial coordinate synchronization. This method significantly improves the realism and transmission robustness of live broadcast images through multi-dimensional dynamic optimization.

[0065] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0066] Figure 1 The present invention provides a flowchart of a method for intelligently sensing and associating multimedia content of a live sports event. Figure 1 As shown, the method includes:

[0067] 101. Obtain original video data;

[0068] In this step, the original video data refers to the unprocessed original image stream captured by the live sports event camera (such as an H.264 encoded video stream without color correction).

[0069] In an embodiment of the present application, first, raw video data, including image streams from different angles, is collected using multiple camera devices deployed at a sports event (e.g., fixed cameras, jibs, or drones). Secondly, the video signals output by each camera (e.g., HDMI or SDI interfaces) are transmitted to a video processing server for frame synchronization and storage based on the device number and timestamp. Next, a preliminary verification (e.g., resolution, frame rate, color depth) of the raw video data is performed to ensure image integrity. Finally, the verified raw video data is sent to a color adjustment module to provide basic input for subsequently eliminating color differences among multiple cameras.

[0070] 102. Generate color adjustment parameters that match the color range distribution of the current sports image based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data. The purpose of the conversion is to eliminate color jumps between images at multiple camera positions.

[0071] In this step, light reflection characteristics refer to the distribution of light reflection intensity at the camera's location due to the venue's material (e.g., lawn, plastic) (e.g., specular reflection accounts for 30%). Color temperature variation refers to the fluctuation in the color temperature of the venue's lighting source over time or weather (e.g., 6500K on a sunny day to 5500K on a cloudy day). Color jump refers to the visual discontinuity caused by sudden color changes when switching between multiple cameras (e.g., camera A appears bluish, camera B appears yellowish). The target video data refers to the unified, standard video stream that eliminates color differences between multiple cameras after color space conversion.

[0072] 103. Adjust the priority of coding block division according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, and adaptively compensate the coding quality parameters based on the brightness distribution of the bright area in the target video data;

[0073] In this step, instantaneous motion speed refers to the rate of pixel change caused by the displacement of a moving object (such as a football) in a single frame (such as 120 pixels per second). Texture complexity refers to the density of detailed areas in the picture (such as the texture complexity of the audience seats is three times higher than that of the grass). Coding block prioritization refers to the priority processing of specific area blocks based on motion characteristics during video encoding (such as 16x16 blocks over 8x8 blocks). Brightness distribution refers to the statistical proportion of pixels at different brightness levels (such as 0-255) in the picture (such as the highlight area accounts for 15% of the picture). Adaptive compensation refers to the dynamic adjustment of the encoding bit rate distribution based on brightness (such as a 20% increase in the bit rate of the highlight area).

[0074] 104. Perform bidirectional data exchange based on the color adjustment parameter and the compensated encoding quality parameter, so that the color component weight ratio in the color adjustment parameter affects the code stream distribution ratio of the color component in the video encoding process;

[0075] In this step, the color component weight ratio refers to the proportion of influence of the red, green, and blue channels in color adjustment (e.g., R: 35%, G: 50%, B: 15%). The bitrate allocation ratio refers to the proportion of data bandwidth allocated to each color component during video encoding (e.g., 70% for the Y component in YUV format). Bidirectional data interaction refers to the adjustment mechanism in which color adjustment parameters and encoding parameters are mutually feedbacked.

[0076] 105. Based on the target video data and the two-way data interaction result, construct a live sports event picture, and associate the live sports event picture with the audio time series data, motion trajectory data, and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence.

[0077] In this step, motion trajectory data refers to the coordinate sequence of the athlete / object's motion path obtained through computer vision tracking (e.g., the three-dimensional coordinates of a soccer ball sampled 30 times per second). Live content transmission sequence refers to the encapsulated standardized transmission unit (e.g., a TS stream containing video frames, synchronized audio, and metadata).

[0078] In summary, steps 101 to 105 realize a systematic technical upgrade of the intelligent processing of multimedia content in live sports events. By dynamically generating color adjustment parameters and encoding quality compensation mechanisms, this method optimizes video encoding efficiency while improving the color reproduction of the picture, and solves the problem of color cast caused by changes in ambient light in traditional live broadcasts. Through two-way data interaction, closed-loop control of color component weights and code stream distribution is achieved, so that color space conversion and encoding compression are synergistically optimized, significantly improving the transmission stability of video streams in complex sports scenes. In addition, by constructing a transmission sequence through multi-dimensional data association, the underlying data support is provided for subsequent multi-perspective switching and content interaction, forming a full-process technical closed loop from picture optimization to content integration.

[0079] In some embodiments, the bidirectional data exchange according to the color adjustment parameter and the compensated encoding quality parameter in step 104, such that the color component weight ratio in the color adjustment parameter affects the bitstream distribution ratio of the color component during the video encoding process, includes:

[0080] 201. Extracting weight ratio values ​​of red, green, and blue color components from the color adjustment parameter, and extracting compression rate limit values ​​corresponding to different coding block priority levels from the compensated coding quality parameter;

[0081] In step 201, the compression rate limit value refers to the maximum compression ratio threshold allowed by different coding block division modes (such as CTU64 or CTU32) (such as CTU64 compression rate ≤ 85%).

[0082] In this embodiment of the present application, first, the weighted ratios of the red, green, and blue color components are parsed from the color adjustment parameters generated in step 102 (e.g., red accounts for 35%, green accounts for 45%, and blue accounts for 20%). These values ​​are directly extracted using the channel gain parameters during color space conversion. Secondly, the compression rate limits corresponding to different coding block priorities (e.g., 16x16 macroblocks have a high priority) are read from the compensated coding quality parameters generated in step 103 (e.g., a 70% upper limit for the compression rate of high-speed motion blocks). Next, the weighted ratios of the red, green, and blue components are superimposed on the compression rate limits for different coding block priorities according to the device's pre-stored weight ratios (e.g., 60% for color weight and 40% for compression rate). Finally, an adaptation relationship parameter is generated that reflects the correlation between color component weights and compression rates (e.g., the compression rate limit for the red component in a macroblock is increased to 65%), providing a basis for subsequent bitstream allocation.

[0083] 202. Superimpose the weight ratio values ​​of the red, green, and blue color components with the compression rate limit values ​​of the corresponding coding block priority according to the preset weight ratio to generate an adaptation relationship parameter between the color component weight and the compression rate;

[0084] In step 202, the adaptation parameter refers to the combined value of the color component weight and the encoding compression limit (e.g., red component weight 0.4 × compression limit 0.8 = 0.32 adaptation value). The preset weight ratio refers to the manually set mixing ratio of color and compression elements (e.g., color weight 60% and compression limit 40%).

[0085] In this embodiment of the present application, first, the weight ratio values ​​of the red, green, and blue color components extracted in step 201 (e.g., a red weight of 30%) and the compression rate limit value (e.g., a corresponding coding block compression rate upper limit of 60%) are weighted and calculated according to a preset superposition rule (e.g., a color weight of 70% and a compression rate of 30%). Secondly, a linear superposition algorithm (e.g., weight ratio × 0.7 + compression rate × 0.3) is used to generate an adaptation relationship value for each color component and coding block priority (e.g., the final adaptation value for the red component is 30% × 0.7 + 60% × 0.3 = 39%). Next, the adaptation values ​​of the red, green, and blue components are combined to form an adaptation relationship parameter for the color component weight and compression rate (e.g., 39% for red, 45% for green, and 16% for blue). Finally, the adaptation relationship parameter is stored in a dynamic configuration table for use in the subsequent formulation of bitstream allocation rules.

[0086] 203. According to the numerical values ​​of the adaptation relationship parameters, a positive correlation between the color component weight ratio values ​​and the compression rate limit values ​​and the bitstream distribution ratio is set to generate a bitstream distribution ratio mapping rule;

[0087] In step 203, the positive correlation refers to a linear relationship in which the corresponding bitrate allocation ratio increases with increasing color component weight (e.g., for every 1% increase in red weight, the bitrate increases by 0.8%). The bitrate allocation ratio mapping rule refers to a conversion formula for automatically generating the bitrate allocation ratio based on the adaptation parameter value (e.g., an adaptation parameter of 0-50 corresponds to a bitrate ratio of 30-70%).

[0088] In the embodiment of the present application, first, based on the numerical value of the adaptation relationship parameter generated in step 202 (e.g., the red component adaptation value is 39%), a positive association rule is set between the color weight ratio and the compression rate limit on the bitstream allocation (e.g., for every 1% increase in the adaptation value, the bitstream allocation ratio increases by 0.5%). Secondly, a mapping relationship between different color components and bitstream proportions is established through piecewise function or threshold judgment (e.g., when the adaptation value is lower than 30%, the bitstream allocation ratio is fixed to the minimum value). Next, the mapping rules are written into the bitstream allocation ratio mapping rule table (e.g., the bitstream allocation ratio corresponding to the red component is capped at 45%). Finally, based on the mapping rules, bitstream allocation templates are dynamically generated under different coding block priorities to ensure a balance between color fidelity and compression efficiency.

[0089] 204. During the video encoding process, detecting a change in the weight ratio of the red, green, and blue color components in the color adjustment parameter, and adjusting the code stream allocation ratio of each color component according to the code stream allocation ratio mapping rule;

[0090] In step 204 , the weight ratio change refers to the dynamic adjustment of the color components due to the change in lighting conditions (eg, the weight of the blue component increases from 15% to 25% on a cloudy day).

[0091] In an embodiment of the present application, first, when the video encoder is running, the weight ratio changes of the red, green, and blue components of the color adjustment parameters in step 102 are monitored (such as the red weight increases from 30% to 35%). Secondly, according to the stream allocation ratio mapping rule generated in step 203, the stream allocation ratio corresponding to the current color component weight is found (such as a red weight of 35% corresponds to a stream allocation ratio of 40%). Then, the bit rate control module of the encoder (such as a macroblock-based stream distributor) is used to dynamically adjust the bit allocation of each color component (such as the red component stream ratio is increased to 40%, and the green and blue are proportionally reduced). Finally, the adjusted stream allocation ratio is synchronized to the encoder configuration to ensure that the stream distribution of the current frame matches the color characteristics.

[0092] 205. Compare the adjusted bitrate allocation ratio with the brightness distribution data of the bright area in the compensated coding quality parameter frame by frame to generate a bitrate allocation deviation feedback signal;

[0093] In step 205, the bitrate allocation deviation feedback signal is a quantitative indicator of the difference between the actual bitrate allocation and the expected value (e.g., the red component bitrate deviation is -3.2%). Frame-by-frame comparison refers to pixel-level verification of the brightness distribution and bitrate allocation of each frame.

[0094] In the embodiment of the present application, first, after each frame is encoded, the adjusted bitrate allocation ratio in step 204 (e.g., a 40% red bitrate ratio) is compared with the bright area brightness distribution data (e.g., a 20% highlight area ratio) in the compensated encoding quality parameters in step 103. Secondly, a difference calculation is performed to determine whether the bitrate allocation matches the brightness distribution requirement (e.g., if the highlight area bitrate requirement is 25%, an actual allocation of 20% results in a 5% deviation). Next, a bitrate allocation deviation feedback signal (e.g., a signal indicating insufficient bitrate in the highlight area of ​​the red component) is generated based on the deviation direction (e.g., insufficient or excessive bitrate). Finally, the feedback signal is marked in the metadata of the corresponding frame to provide a basis for reverse correction.

[0095] 206. According to the bit stream allocation deviation feedback signal, reversely correct the weight ratio of the red, green, and blue color components in the color adjustment parameter to form a two-way data interaction closed loop of the color component weight and the bit stream allocation ratio, and generate a two-way data exchange result.

[0096] In step 206, reverse correction refers to adjusting the color weight inversely according to the bitrate deviation (e.g., if the bitrate is over-matched, the corresponding color weight is reduced by 2%). The two-way data interaction closed loop refers to a cyclic optimization system for color adjustment and bitrate allocation.

[0097] In this embodiment, first, based on the bitrate distribution deviation feedback signal from step 205 (e.g., insufficient bitrate in the highlight area of ​​the red component), the weighting ratios of the red, green, and blue color components in the color adjustment parameters of step 102 are reversed (e.g., increasing the red weight from 35% to 38%). Secondly, a closed-loop control algorithm (e.g., PID regulation) is used to gradually correct the weighting ratios until the bitrate distribution deviation approaches zero. The corrected color weighting parameters are then re-entered into the process of steps 201 to 205, forming a bidirectional data exchange loop between color component weights and bitrate distribution ratios. Finally, a bidirectional data exchange result containing the dynamic balancing result is output (e.g., red weight stabilized at 36% and bitrate distribution deviation less than 1%), ensuring long-term stability of color consistency, compression efficiency, and image quality.

[0098] In summary, steps 201 to 206 achieve dynamic collaborative optimization of color component and bitstream allocation during video encoding. By establishing an adaptive relationship between the color component weight ratio and the compression rate, this method innovatively links color features with encoding parameters in a bidirectional manner, resolving the contradiction between compression efficiency and image quality loss caused by the rigid bitstream allocation of color components in traditional encoding. The closed-loop feedback mechanism detects the bitstream allocation deviation and reversely corrects the color parameters, thereby enhancing the adaptability of the encoding process to complex lighting changes, effectively avoiding color distortion or blurred details caused by unbalanced bitrate allocation in high-dynamic scenes, and significantly improving the color fidelity and subjective visual quality of the compressed video.

[0099] In some embodiments, the step 102 generates color adjustment parameters that match the color range distribution of the current motion picture based on the light reflection characteristics of different camera positions in the live sports event scene and the changes in the color temperature of the field lighting, and performs nonlinear conversion on the color space of the original video data using the color adjustment parameters to obtain the target video data, including:

[0100] 301. Based on the light reflection characteristics of different camera positions in the live sports event scene, extract the light intensity distribution curve and reflection angle range of each camera position, and simultaneously collect the color temperature values ​​of the venue lighting color temperature changes to generate the current light reflection and color temperature parameter group for each camera position;

[0101] In step 301, the light intensity distribution curve refers to the gradient of light intensity across the camera's field of view (e.g., 2000 lux in the center, 800 lux at the edges). The reflection angle range refers to the primary angle of incidence of light reflected from the surface (e.g., 30°-60° specular reflection angle). The color temperature parameter set refers to a data set containing a color temperature value (e.g., 5600K) and a color deviation value (ΔE = 2.3).

[0102] In an embodiment of the present application, first, for different camera positions in a live sports event scene (such as the main stand, the goal side, and the high-altitude jib), the light intensity distribution curve (such as the curve showing the change in the reflection intensity of a metal guardrail with angle) and the reflection angle range (such as the maximum incident angle of reflection being 30°-60°) of the area in which the camera is located are collected by a light sensor installed on the camera. Secondly, the color temperature value (such as 5600K on a sunny day and 6500K on a cloudy day) is obtained through the color temperature sensor of the venue lighting system, and the lighting parameters corresponding to each camera position are recorded. Next, the light intensity distribution curve, reflection angle range, and color temperature value are integrated into the current light reflection and color temperature parameter group according to the camera number (such as the main stand camera position parameter group includes reflection curve A, angle range 30°-50°, and color temperature 5600K). Finally, the parameter group is sent to the color adjustment module to provide an environmental data basis for subsequent color adaptation.

[0103] 302. Extracting a distribution histogram of the red, green, and blue color components of the current motion picture from the original video data, and calculating adaptation weights of the red, green, and blue color components based on the reflection angle range and color temperature value in the light reflection and color temperature parameter group;

[0104] In step 302, the distribution histogram refers to the pixel distribution histogram of the red, green, and blue color channels in the statistical image (for example, the red channel has a peak at grayscale 128). The adaptation weight refers to the adjustment ratio of the color component calculated based on the lighting conditions (for example, the green component weight is increased to 55%).

[0105] In an embodiment of the present application, first, the distribution histograms of the red, green, and blue color components are extracted from the current frame of the original video data (e.g., red is concentrated in high-brightness areas, and green is in medium- and low-brightness areas). Secondly, based on the light reflection generated in step 301 and the reflection angle range in the color temperature parameter group (e.g., a reflection angle of 50°-70°), the weight of the influence of the reflection on the color component is calculated (e.g., the red channel needs to reduce the gain by 20% at high reflection angles). Then, combined with the color temperature value (e.g., 6500K is cooler), the histograms of the red, green, and blue components are proportionally corrected through a color correction algorithm (e.g., the gain of the blue component is increased by 5% to neutralize the cool tones) through a color correction algorithm (e.g., white balance adjustment). Finally, an adaptation weight is generated for each color component (e.g., red weight 0.8, green 1.0, blue 1.2) for subsequent generation of color adjustment parameters.

[0106] 303. Scaling the distribution histograms of the red, green, and blue color components based on the adaptation weights to generate color adjustment parameters that match the color range distribution of the current motion picture, the color adjustment parameters including nonlinear offset coefficients and gain coefficients of the red, green, and blue color components;

[0107] In step 303, the nonlinear shift coefficient refers to the curved hue shift applied to a specific color range (e.g., a deep red region) (e.g., increasing the saturation of the R channel by 5% in the dark regions). The gain coefficient refers to the linear amplification factor of the color component (e.g., a 1.2x overall gain for the blue channel).

[0108] In an embodiment of the present application, first, based on the adaptation weight calculated in step 302 (such as a red weight of 0.8), the distribution histograms of the red, green, and blue color components are scaled (such as the red histogram is compressed to 80% of the original range as a whole). Secondly, a nonlinear curve fitting algorithm (such as a gamma correction curve) is used to generate a nonlinear offset coefficient (such as an offset of +0.1 in the low brightness area of ​​the red channel) and a gain coefficient (such as a global gain of ×1.1 for the blue channel) for each color component based on the scaled histogram. Then, the offset coefficient and the gain coefficient are integrated into color adjustment parameters (such as a red offset coefficient table and a blue gain mapping table) according to the color channel. Finally, the color adjustment parameters are written into a configuration file to guide the color space conversion of the original video data.

[0109] 304. Perform piecewise nonlinear mapping on the red, green, and blue color components of the original video data using the nonlinear offset coefficient and gain coefficient in the color adjustment parameters, and convert the original color space into a target color space to generate target video data.

[0110] In step 304, piecewise nonlinear mapping involves dividing the color space into multiple intervals (e.g., 0-64, 65-192, 193-255) and applying different conversion curves to each interval. The target color space is a color space that complies with broadcast standards after conversion (e.g., the BT.709 color gamut).

[0111] In an embodiment of the present application, first, the nonlinear offset coefficient and gain coefficient (such as red offset coefficient +0.1, blue gain × 1.1) in the color adjustment parameters generated in step 303 are read. Secondly, each frame of the original video data is segmented according to the brightness range of the red, green, and blue color components (such as dividing red into three segments: low, medium, and high), and the corresponding nonlinear offset and gain are applied respectively (such as increasing the low-brightness red pixel value by 10%). Then, the adjusted RGB components are mapped to the target color space (such as the BT.709 color gamut of the broadcast standard) through a color space conversion algorithm (such as RGB to YUV matrix transformation). Finally, the target video data with consistent color and matching the lighting environment is output to eliminate the color temperature differences and reflection interference between multi-camera images.

[0112] In summary, steps 301 to 304 achieve a breakthrough in adaptive color correction technology based on environmental perception. By precisely extracting the light reflection characteristics and color temperature parameters at the camera's location, this method establishes a mapping model between scene characteristics and color adjustment parameters, overcoming the reliance of traditional color correction techniques on manually preset parameters. The nonlinear color space conversion mechanism, through a segmented mapping strategy, accurately corrects color deviations under varying lighting conditions while preserving natural image transitions. This particularly addresses the industry challenge of white balance drift in multi-light source environments in sports stadiums, providing viewers with a visual experience closer to the real-life experience.

[0113] In some embodiments, performing piecewise nonlinear mapping on the red, green, and blue color components of the original video data using the nonlinear offset coefficients and gain coefficients in the color adjustment parameters in step 304 to convert the original color space into the target color space to generate the target video data includes:

[0114] 401. Divide the value range of each color component into a low value interval, a medium value interval, and a high value interval based on the distribution histogram value range of the red, green, and blue color components in the original video data, with the division boundaries aligned with the peak and valley values ​​of the histogram;

[0115] In step 401, the low value interval refers to the portion of the color component distribution histogram with lower values ​​(e.g., 0-85 for red). The medium value interval refers to the portion of the color component distribution histogram with intermediate values ​​(e.g., 86-170 for green). The high value interval refers to the portion of the color component distribution histogram with higher values ​​(e.g., 171-255 for blue). Peak and valley alignment refers to matching the peaks (highest frequency values) and valleys (lowest frequency values) of the histogram when dividing the interval boundaries (e.g., if the peak is at 128, the boundaries are set to 100 and 150).

[0116] In an embodiment of the present application, first, the numerical range of each color component is extracted from the distribution histogram of the red, green, and blue color components generated in step 302 (such as the numerical range of the red component 0-255). Secondly, the boundaries of the low, medium, and high numerical intervals of each component are determined by a histogram analysis algorithm (such as peak detection and valley location): the low numerical interval is the range before the first valley value on the left side of the histogram (such as red 0-50), the medium numerical interval is the range between the two valley values ​​(such as red 51-200), and the high numerical interval is the range after the valley value on the right (such as red 201-255). Then, the interval boundaries are dynamically adjusted to align with the peak position of the histogram (such as the main red peak is located in the center of the medium numerical interval). Finally, the three divided interval boundaries are written into the configuration file to provide a basic basis for segmented mapping.

[0117] 402. Bind the low value interval, the middle value interval, and the high value interval of each color component to the nonlinear offset coefficient and the gain coefficient, respectively, and set a segmented mapping rule in which the low value interval is bound to the high gain coefficient and the positive offset coefficient, and the high value interval is bound to the low gain coefficient and the negative offset coefficient;

[0118] In step 402, the segmented mapping rule indicates a conversion strategy that sets different offset and gain coefficients for different value intervals (e.g., a gain of 1.5 and an offset of +10 in the low range). A positive offset coefficient increases the value of a color component (e.g., a red component of +15). A negative offset coefficient decreases the value of a color component (e.g., a blue component of -8).

[0119] In the embodiment of the present application, first, the parameters corresponding to each color component are extracted from the nonlinear offset coefficient and gain coefficient generated in step 303 (e.g., a gain coefficient of 1.2 and an offset coefficient of +10 for the low red interval). Secondly, according to the preset rules, the low numerical interval is bound to a high gain coefficient and a positive offset coefficient (e.g., a gain of 1.2 and an offset of +10 for the low red interval), the medium numerical interval is bound to a medium gain and a zero offset (e.g., a gain of 1.0 and an offset of 0 for the middle red interval), and the high numerical interval is bound to a low gain coefficient and a negative offset coefficient (e.g., a gain of 0.8 and an offset of -5 for the high red interval). Next, the three intervals of each color component and the corresponding parameters are combined into a segmented mapping rule table (e.g., a red interval of 0-50 → a gain of 1.2 / an offset of +10). Finally, the mapping rule table is loaded into the color conversion module for pixel-level processing.

[0120] 403. For the values ​​of the red, green, and blue color components of each pixel in the original video data, invoke the bound segment mapping rule according to the value interval to which the values ​​belong, and map the target color component values;

[0121] In step 403, the segmented mapping rule call refers to selecting the corresponding conversion parameters based on the pixel color value interval (e.g., a red value of 120 calls for a gain of 1.2 and an offset of +5 in the interval). The target color component value refers to the new value of the color component after the mapping (e.g., the original green value of 180 is mapped to 165).

[0122] In the embodiment of the present application, first, the red, green, and blue color component values ​​of each pixel in the original video data are read (for example, the red value of a pixel is 45, the green value is 120, and the blue value is 210). Secondly, the interval to which each component belongs is determined according to the numerical interval divided in step 401 (for example, red 45 belongs to the low interval, green 120 belongs to the middle interval, and blue 210 belongs to the high interval). Next, the nonlinear offset coefficient and gain coefficient of the corresponding interval are searched from the segmented mapping rule table in step 402 (for example, the gain of the red low interval is 1.2, and the offset is +10). Then, the original value is substituted into the mapping formula (for example, target value = original value × gain coefficient + offset coefficient) to calculate the target color component value (for example, red target value = 45 × 1.2 + 10 = 64). Finally, the mapped red, green, and blue component values ​​are temporarily stored in the intermediate buffer, waiting for overlay processing.

[0123] 404 : Superimpose the mapped target color component values ​​according to the scaling rule of the distribution histogram to generate a superimposed value and use the superimposed value as the color value of each pixel in the target video data to generate the target video data.

[0124] In step 404, the superimposed value refers to the synthesis result of each color component after segment mapping and scaling.

[0125] In the embodiment of the present application, first, the target color component value of each pixel is obtained from the intermediate buffer in step 403 (e.g., red 64, green 120×1.0=120, blue 210×0.8-5=163). Secondly, according to the scaling rule of the distribution histogram in step 303 (e.g., the overall scaling of the green component to 90% range), the mapped values ​​are superimposed and adjusted (e.g., green value 120×0.9=108). Next, the adjusted results of the red, green, and blue components are combined into the final pixel value (e.g., RGB(64,108,163)) in a color space format (e.g., RGB). Finally, target video data with consistent color and optimized dynamic range is generated frame by frame, completing the elimination and unification of color differences in multi-camera images.

[0126] In summary, steps 401 to 404 achieve refined nonlinear control of color space conversion. By dividing the color components into different numerical intervals and binding differentiated mapping rules, this method innovatively solves the phenomenon of dark detail loss or highlight overflow caused by traditional linear conversion. The segmented processing based on histogram features effectively balances the color expression of each brightness interval, enhances the dark color level through high gain coefficients, and suppresses overexposure in highlight areas through negative offset coefficients. While maintaining the color coordination of the overall picture, it enhances the visual expression of local areas, greatly improving the image quality presentation effect in night events or backlit scenes.

[0127] In some embodiments, the step 103 includes adjusting the priority of coding block division according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, and adaptively compensating the coding quality parameters in combination with the brightness distribution of the bright area in the target video data, including:

[0128] 501. Extract the instantaneous motion speed of the moving object from the live sports event footage, calculate the displacement difference between adjacent video frames, and simultaneously extract the edge sharpness change of the image texture complexity to generate a joint evaluation parameter of motion speed and texture complexity;

[0129] In step 501, the displacement difference refers to the distance in pixels that a moving object moves between adjacent frames (e.g., a soccer ball moves 25 pixels between two frames). The edge sharpness change refers to the magnitude of the gradient change at the edge of an object in the image (e.g., the edge sharpness change at the audience seats is 0.3). The joint evaluation parameter is an evaluation metric generated by combining motion speed and texture complexity.

[0130] In an embodiment of the present application, the instantaneous motion speed of a moving object (e.g., an athlete or a soccer ball) is first calculated from live sports footage using an optical flow method or an inter-frame difference algorithm. This is specifically the pixel displacement difference between adjacent video frames (e.g., a soccer ball moves 30 pixels between two frames). Next, an edge detection algorithm (e.g., the Sobel operator) is used to extract the texture complexity of a local area of ​​the image. Texture complexity is quantified by calculating the gradient change in edge sharpness (e.g., the gradient value in a grass area fluctuates widely). Next, the displacement difference and the edge sharpness change are multiplied by a preset ratio (e.g., the displacement difference weights 60% and the texture complexity weights 40%) to generate a joint evaluation parameter for motion speed and texture complexity (e.g., the evaluation value for a particular area is 30 × 0.6 + 80 × 0.4 = 50). Finally, the joint evaluation parameter is stored in a grid format by image region, providing a basis for prioritizing and adjusting coding blocks.

[0131] 502. Adjust the priority of coding block division according to the product relationship between the displacement difference and the edge sharpness change in the joint evaluation parameter, and generate a coding block priority distribution map for high motion areas and high texture areas;

[0132] In step 502, the coding block priority distribution map refers to a spatial distribution map that labels the processing priorities of coding blocks based on motion and texture characteristics (e.g., high-motion areas are marked as priority 1). High-motion areas refer to areas of the image where object displacement exceeds a threshold (e.g., areas where athletes are running). High-texture areas refer to areas of the image where the change in edge sharpness exceeds a threshold (e.g., the intersection of the lawn and the sideline).

[0133] In an embodiment of the present application, first, according to the joint evaluation parameter generated in step 501 (such as a certain area value of 50), the product relationship between the displacement difference and the edge sharpness change is analyzed (such as the higher the product value, the greater the influence of the superposition of motion and texture complexity). Secondly, the coding block division priority rule is set: when the product value exceeds the high threshold (such as> 60), it is marked as a high motion area (prioritized to 16x16 macroblocks), and when the product value is medium (such as 30-60), it is marked as a high texture area (prioritized to 8x8 sub-blocks). Then, a coding block priority distribution map of the entire screen is generated through the regional priority mapping algorithm (such as high motion areas are marked in red and high texture areas are marked in yellow). Finally, the distribution map is bound to the encoder configuration to guide the subsequent coding block division strategy.

[0134] 503. Extract the brightness distribution of bright areas from the target video data, calculate the area ratio of areas whose brightness values ​​exceed a preset highlight threshold, and generate a brightness distribution weight coefficient;

[0135] In step 503, the highlight threshold refers to the brightness threshold for determining bright areas (e.g., brightness ≥ 200). The area ratio refers to the percentage of the highlight area to the total image area. The brightness distribution weight coefficient refers to the compensation weight dynamically generated based on the area of ​​the highlight area.

[0136] In the embodiment of the present application, first, the brightness channel data is extracted from the target video data generated in step 404, and the bright areas (such as pixels with brightness values ​​exceeding 200) are determined through threshold segmentation. Secondly, the area ratio of the bright area in the picture is calculated (such as the highlight area accounts for 15% of the total area of ​​the picture). Then, according to the preset brightness weight curve (such as the weight coefficient increases by 10% for every 5% increase in area ratio), the brightness distribution weight coefficient is generated (such as 15% corresponds to a weight coefficient of 30%). Finally, the weight coefficient is written into the configuration file according to the picture area distribution, providing a brightness correlation basis for encoding quality parameter compensation.

[0137] 504. Perform weighted compensation on the compression strength and detail retention ratio in the coding quality parameters according to the priority levels of different areas in the coding block priority distribution map and in combination with the brightness distribution weight coefficient to generate adaptively compensated coding quality parameters.

[0138] In step 504, compression strength refers to the strength of data compression during encoding. Detail retention ratio refers to the degree of image detail retained after encoding. Weighted compensation refers to adjusting compression parameters based on priority and brightness weight.

[0139] In an embodiment of the present application, first, the coding block priority distribution map of step 502 is read to obtain the priority levels of different areas (such as the high motion area priority level is 3 and the normal area priority level is 1). Secondly, combined with the brightness distribution weight coefficient of step 503 (such as the brightness weight of a certain area is 30%), the compression strength (such as the quantization step size) and the detail retention ratio (such as the DCT coefficient retention rate) in the coding quality parameters are weightedly calculated: the higher the priority level and the larger the brightness weight, the lower the compression strength (the step size is reduced) and the detail retention ratio is increased. Then, the adaptively compensated coding quality parameters are generated through the dynamic parameter compensation algorithm (such as the quantization step size of the high motion highlight area is reduced from 30 to 25). Finally, the compensation parameters are synchronized to the encoder to achieve bit rate allocation for joint optimization of motion, texture and brightness.

[0140] In summary, steps 501 to 504 achieve intelligent optimization of video encoding parameters in motion scenarios. By jointly evaluating motion speed and texture complexity parameters, this method establishes a dynamic priority adjustment mechanism, addressing the motion blur and texture distortion issues caused by traditional fixed coding block partitioning. Combining a compensation strategy based on luminance distribution characteristics and introducing spatial weighting coefficients into the encoding quality parameters, this method achieves a balanced optimization between detail preservation in high-motion areas and noise control in brightly lit areas. This significantly improves image clarity in intense game scenarios, such as rapid offense-defense transitions, providing technical support for viewers to capture crucial game moments.

[0141] In some embodiments, step 105 includes constructing a live sports event image based on the target video data and the bidirectional data interaction result, and associating the live sports event image with the audio time series data, motion trajectory data, and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence, including:

[0142] 601. Extracting a video frame sequence after color space conversion from the target video data, and extracting a color component weight ratio, a bitrate allocation ratio parameter, and an encoding quality compensation parameter from a bidirectional data interaction result, and combining them with the video frame sequence to generate a live picture construction parameter group;

[0143] In step 601 , the live broadcast image construction parameter group refers to a set of color adjustment, bitrate allocation, and encoding compensation parameters (eg, {R weight 0.4, bitrate allocation Y: 70%}).

[0144] In an embodiment of the present application, first, a video frame sequence after color space conversion (such as YUV format data of 25 frames per second) is extracted from the target video data generated in step 404. Secondly, the color component weight ratio (such as red accounting for 35%), the stream allocation ratio parameter (such as Y component stream accounting for 70%) and the encoding quality compensation parameter (such as the quantization step size of the highlight area is reduced by 20%) are read from the two-way data interaction result of step 206. Then, the above parameters are associated with the metadata of the video frame sequence (such as frame number, timestamp) to form a live screen construction parameter group including color adjustment, stream allocation and image quality compensation (such as the parameter group includes frame 1 red weight of 35%, stream allocation ratio table, quantization step size table). Finally, the parameter group is loaded into the live screen synthesis module to provide dynamic configuration for screen rendering.

[0145] 602. Adjust the numerical distribution of red, green, and blue color components in the video frame according to the color component weight ratio in the live image construction parameter group, and dynamically allocate the bitstream resources of each frame in combination with the bitstream allocation ratio parameter to generate a live sports event image including coding quality compensation.

[0146] In step 602, dynamic allocation of code stream resources refers to allocating different bit rates based on the characteristics of the regions within the frame (e.g., allocating 30% bit rate to high motion regions). Coding quality compensation refers to reducing compression distortion by adjusting parameters (e.g., increasing the bit rate of highlight regions by 10%).

[0147] In an embodiment of the present application, first, according to the weight ratio of the color components in the live picture construction parameter group (such as 35% for red), the red, green, and blue color components of the video frame are numerically adjusted (such as the overall pixel value of the red channel is increased by 5%). Secondly, in combination with the bit stream allocation ratio parameter (such as 70% bit stream allocated to the Y component), the bit resources of each frame are dynamically allocated through the encoder's bit rate control module (such as allocating more bit stream to the Y component in high-motion areas). Then, based on the encoding quality compensation parameters (such as reducing the quantization step by 20%), the picture details are enhanced (such as reducing the mosaic in the highlight area). Finally, a live broadcast picture of a sports event with consistent color, optimized bit stream allocation, and compensated image quality is generated (such as no ghosting when athletes move at high speed).

[0148] 603. Align each time node of the audio time series data in the live stream with the video frame timestamp of the live sports event image frame by frame to generate synchronization timestamp parameters for the audio and video frames;

[0149] In step 603, the synchronization timestamp parameter refers to the precise mark of the time alignment of the audio and video frames (eg, the 5th second of the audio corresponds to the 120th frame of the video).

[0150] In an embodiment of the present application, first, audio time series data (such as the sampling point sequence of the commentator's voice) is extracted from the audio acquisition device of the live stream, and the time node of each audio segment is recorded (such as the voice at the 10th second corresponds to frame 250). Secondly, the timestamp is extracted from the video frame metadata of the live sports event screen (such as the timestamp of frame 250 is 00:00:10:00). Then, the audio time node is matched with the video frame timestamp frame by frame through a timestamp alignment algorithm (such as a synchronization mechanism based on PTS / DTS) (such as the 10th second of the voice is bound to frame 250). Finally, a synchronization timestamp parameter containing the absolute time relationship between audio and video is generated (such as the synchronization mark of frame 250 is AUDIO_SYNC_250) to ensure that there is no delay in the playback of audio and video.

[0151] 604. Matching the position coordinates of the athletes in the motion trajectory data in the live stream with the corresponding view angles in the multi-camera position spatial coordinates to obtain spatial anchor point parameters of the motion trajectory and the camera view angle in each video frame;

[0152] In step 604, the spatial anchor point parameter refers to the spatial coordinate mapping relationship between the motion trajectory and the camera perspective (eg, the athlete coordinate (10, 20) corresponds to the center of the camera C3 screen).

[0153] In an embodiment of the present application, first, the position coordinates of the athlete are extracted from the motion trajectory data of the live stream (such as the GPS coordinates of the football X=120, Y=80). Secondly, based on the spatial coordinates of the multi-camera positions (such as the left camera coordinates X=0, Y=0), the position of the motion trajectory in the corresponding camera screen is calculated through a perspective mapping algorithm (such as the projection transformation of 3D coordinates to 2D screen) (such as the coordinates of the football in the left camera screen are X=300, Y=200). Then, the trajectory coordinates are bound to the camera perspective to generate spatial anchor point parameters (such as the anchor point mark of frame 250 is CAM1_ANCHOR(300,200)). Finally, the anchor point parameters are written into the video frame metadata to provide a spatial alignment basis for multi-perspective switching.

[0154] 605. Data association and encapsulation of the sports event live broadcast images, synchronization timestamp parameters, and spatial anchor point parameters are performed according to a preset transmission format to generate a live content transmission sequence.

[0155] In step 605, the preset transmission format refers to a standardized protocol for encapsulating live data (such as MPEG-TS format).

[0156] In an embodiment of the present application, first, the live sports event picture generated in step 602, the synchronization timestamp parameters in step 603, and the spatial anchor point parameters in step 604 are data-associated according to the transmission format specification (such as the MPEG-TS stream format) (such as binding the synchronization mark and anchor point coordinates to frame 250). Secondly, the picture, audio, trajectory, and coordinate data are packaged into a unified data block (such as a TS segment containing a video PES packet, an audio PES packet, and a metadata descriptor) through an encapsulation protocol (such as FLV or HLS segmentation). Then, a synchronization mark (such as a PCR clock reference) and a spatial anchor point index (such as the CAM1_ANCHOR field) are inserted into the data block. Finally, a complete multi-data fused live content transmission sequence (such as a streaming media file containing audio and video synchronization and multi-view anchor points) is generated and pushed to the CDN node for multi-terminal distribution.

[0157] In summary, steps 601 to 605 achieve deep correlation and synchronous transmission of multi-source heterogeneous data. By constructing a comprehensive transmission sequence that includes visual parameters, encoding strategies, and spatiotemporal information, this method overcomes the technical bottleneck of isolated transmission of audio and video data and motion information in traditional live broadcasts. Dynamic bitrate allocation and timestamp alignment mechanisms ensure precise synchronization of multi-dimensional data, while the introduction of spatial anchor point parameters establishes a mapping relationship between motion trajectories and multi-camera perspectives, providing a data foundation for the intelligent broadcast system to automatically select the optimal shooting perspective, significantly enhancing the sense of presence and information richness of live event broadcasts.

[0158] In some embodiments, as described in step 604, the spatial anchor point parameters of the motion trajectory and the camera perspective in each video frame are obtained by matching the position coordinates of the athlete in the motion trajectory data in the live stream with the corresponding perspectives in the spatial coordinates of the multiple camera positions, including:

[0159] 701. Extracting the position coordinates of the athlete corresponding to the timestamp of each video frame from the motion trajectory data in the live stream, and generating a sequence of the athlete's position coordinates based on the position coordinates;

[0160] In step 701 , the player position coordinate sequence refers to player position data recorded in time sequence (eg, 30 (x, y, z) coordinates per second).

[0161] In an embodiment of the present application, first, the position coordinates corresponding to the timestamp of each video frame are extracted from the motion trajectory data of the live stream (such as the GPS trajectory of a football or the sensor data of an athlete) (such as the timestamp 00:00:04:00 of frame 100 corresponds to the coordinates X=120, Y=80). Secondly, the coordinates are sorted in frame order through a timestamp alignment algorithm to form a sequence of athlete position coordinates that is completely synchronized with the video stream (such as a sequence containing a list of coordinates from frame 1 to frame N). Next, the coordinate data is denoised (such as removing GPS signal drift points) to ensure the continuity of the position sequence. Finally, the processed coordinate sequence is stored as structured data (such as a JSON array) for subsequent camera matching.

[0162] 702. Extract the spatial coordinates of each camera and the coverage of its shooting angle of view from the spatial coordinates of the multiple camera positions;

[0163] In step 702 , the coverage refers to the visible area of ​​the camera's viewing angle in three-dimensional space (eg, a horizontal viewing angle of 60° and a vertical viewing angle of 40°).

[0164] In an embodiment of the present application, first, the spatial coordinates of each camera (such as the left camera position coordinates X=0, Y=0, Z=5) and the coverage range of its shooting angle (such as 60° horizontal angle and 40° vertical angle) are read from a preset multi-camera position spatial coordinate database (such as the site three-dimensional modeling data). Secondly, the viewing angle parameters are converted into the three-dimensional spatial range of the actual shooting area (such as the left camera position covers X-axis 0-200 and Y-axis 0-150) through a viewing angle coverage model (such as cone space calculation). Then, the spatial coordinates of each camera are bound to the coverage range to generate a camera viewing angle configuration table (such as a list of coverage range boundary coordinates of camera position 1). Finally, the configuration table is loaded into the matching module for subsequent association judgment between coordinate points and cameras.

[0165] 703. Perform position matching on each coordinate point in the athlete position coordinate sequence and the coverage range. If the coordinate point is within the coverage range, mark the camera as a matching camera, and obtain a preliminary matching result.

[0166] In step 703 , the preliminary matching result refers to filtering out a list of cameras that may capture the athlete (eg, coordinate (15, 30) matches cameras C1 and C2 ).

[0167] In the embodiment of the present application, first, the coordinate point of a certain frame is extracted from the athlete position coordinate sequence of step 701 (such as the coordinates X=150, Y=90 of frame 250). Secondly, the camera view configuration table of step 702 is traversed to determine one by one whether the coordinate point is within the coverage range of a certain camera (such as checking whether X=150, Y=90 is within the X-axis 0-200 and Y-axis 0-150 range of the left camera). Then, if the coordinate point is within the coverage range of a certain camera (such as the left camera meets the conditions), the camera is marked as a matching camera (such as marking the left camera as the matching device for frame 250). Finally, the matching results are stored in the order of frames as a preliminary matching result table (such as frame 250 matches the left camera, frame 251 has no match, etc.), providing a basis for position mapping.

[0168] 704. Generate position mapping parameters of the relative pixel coordinates of the athlete's position in the camera image based on the relative position of each coordinate point in the preliminary matching result and the spatial coordinates of the matching camera;

[0169] In step 704 , the relative pixel coordinates refer to the two-dimensional position of the athlete in the camera image (eg, (x=320, y=180) in the camera C1 image).

[0170] In the embodiment of the present application, first, a coordinate point and its matching camera are extracted from the preliminary matching result table of step 703 (e.g., the coordinates X=150, Y=90 of frame 250 match the left camera). Secondly, based on the spatial coordinates of the matching camera (e.g., the left camera coordinates X=0, Y=0, Z=5), the relative pixel coordinates of the coordinate point in the camera image are calculated using a coordinate projection algorithm (e.g., 3D to 2D perspective transformation) (e.g., X=150, Y=90 is mapped to pixels X=320, Y=240 in the left camera image). Next, the pixel coordinates are corrected according to the distortion parameters of the camera lens (e.g., fisheye lens correction table) (e.g., X=320 is corrected to X=315). Finally, position mapping parameters containing the correspondence between the coordinate point and the pixel position are generated (e.g., the pixel coordinates X=315, Y=240 of frame 250) and written into the mapping parameter table.

[0171] 705. Bind the position mapping parameters to the spatial coordinates of the matching camera to obtain spatial anchor point parameters.

[0172] In step 705 , binding refers to establishing a fixed association between spatial coordinates and pixel coordinates (eg, camera C2 coordinates (5, 10, 3) correspond to screen pixels (400, 225)).

[0173] In an embodiment of the present application, first, the pixel coordinates of a certain frame (such as X=315, Y=240 of frame 250) and its corresponding matching camera (such as the left camera) are extracted from the position mapping parameters of step 704. Secondly, the pixel coordinates are bound to the spatial coordinates of the matching camera (such as the left camera coordinates X=0, Y=0, Z=5) to form the spatial anchor point parameters under the camera's perspective (such as the anchor point is marked as "left camera_frame 250_X315_Y240"). Then, the anchor point parameters are stored by frame and camera classification through a spatial anchor point indexing algorithm (such as hash table storage). Finally, a spatial anchor point parameter table (such as a CSV file or database record) containing all frame anchor point information is output for dynamic association of live broadcast images with multi-perspective data.

[0174] In summary, steps 701 to 705 achieve intelligent spatial matching of moving targets and camera perspectives. By establishing a dynamic matching model between athlete position coordinates and camera coverage, this method innovatively solves the problem of perspective switching lag in multi-camera live broadcasts. The spatial anchor point parameter generation mechanism based on position coordinates accurately calculates the relative position of moving targets in different camera images, providing a spatial benchmark for perspective prediction and smooth switching of the intelligent directing system, significantly improving the response speed and smoothness of multi-camera collaborative shooting, and enhancing the multi-angle presentation capabilities of key action in the event.

[0175] Figure 2 The present application provides a schematic diagram of a system for intelligently sensing and associating multimedia content of live sports events. Figure 2 As shown, the device includes:

[0176] An acquisition module 21 is used to acquire original video data;

[0177] A conversion module 22 is configured to generate color adjustment parameters that match the color range distribution of the current sports image based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data;

[0178] An adjustment module 23 is configured to adjust the priority of coding block division according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, and to adaptively compensate the coding quality parameters based on the brightness distribution of the bright areas in the target video data;

[0179] An interaction module 24 is configured to perform bidirectional data interaction based on the color adjustment parameters and the compensated encoding quality parameters, so that the color component weight ratio in the color adjustment parameters affects the code stream distribution ratio of the color components during video encoding;

[0180] The construction module 25 is used to construct a live sports event screen based on the target video data and the two-way data interaction results, and associate the live sports event screen with the audio time series data, motion trajectory data and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence.

[0181] Figure 2 The system for intelligently sensing and associating multimedia content of live sports events can be implemented Figure 1 The implementation principles and technical effects of the method for intelligently sensing and associating multimedia content for live sports events described in the illustrated embodiment are not further elaborated. The specific manner in which each module and unit performs operations in the system for intelligently sensing and associating multimedia content for live sports events in the aforementioned embodiment has been described in detail in the relevant embodiments of the method and will not be further elaborated here.

[0182] In one possible design, Figure 2 The system for intelligent perception and association of multimedia content of live sports events in the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0183] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0184] The processing component 32 is used for the above Figure 1 The embodiment provides a method for intelligently sensing and associating multimedia content of a live sports event.

[0185] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0186] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0187] Of course, a computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0188] The input / output interface provides an interface between the processing component and the peripheral interface module, which can be an output device, an input device, etc.

[0189] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0190] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0191] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment provides a method for intelligently sensing and associating multimedia content of a live sports event.

[0192] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0193] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0194] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0195] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for intelligently sensing and associating multimedia content of live sports events, characterized in that: include: Get the original video data; Generate color adjustment parameters that match the color range distribution of the current motion picture based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data; Adjusting the priority of coding block division based on the instantaneous motion speed and image texture complexity of moving objects in live sports events, and adaptively compensating coding quality parameters based on the brightness distribution of bright areas in the target video data; Performing bidirectional data interaction according to the color adjustment parameter and the compensated encoding quality parameter, so that the color component weight ratio in the color adjustment parameter affects the code stream distribution ratio of the color component during the video encoding process; Based on the target video data and the two-way data interaction results, a live sports event picture is constructed, and the live sports event picture is associated with the audio time series data, motion trajectory data, and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence; Performing bidirectional data interaction according to the color adjustment parameter and the compensated encoding quality parameter, so that the color component weight ratio in the color adjustment parameter affects the code stream distribution ratio of the color component in the video encoding process, includes: Extracting weight ratio values ​​of red, green, and blue color components from the color adjustment parameter, and extracting compression rate limit values ​​corresponding to different coding block priority levels from the compensated coding quality parameter; The weight ratio values ​​of the red, green, and blue color components are respectively superimposed with the compression rate limit values ​​of the corresponding coding block priority according to the preset weight ratio to generate the adaptation relationship parameters between the color component weight and the compression rate; According to the numerical value of the adaptation relationship parameter, a positive correlation between the color component weight ratio value and the compression rate limit value and the bitstream distribution ratio is set to generate a bitstream distribution ratio mapping rule; During the video encoding process, detecting a change in the weight ratio of the red, green, and blue color components in the color adjustment parameter, and adjusting the code stream allocation ratio of each color component according to the code stream allocation ratio mapping rule; Comparing the adjusted bitrate allocation ratio with the brightness distribution data of the bright area in the compensated coding quality parameter frame by frame to generate a bitrate allocation deviation feedback signal; According to the code stream allocation deviation feedback signal, the weight ratio of the red, green and blue color components in the color adjustment parameter is reversely corrected to form a two-way data interaction closed loop of the color component weight and the code stream allocation ratio, and generate a two-way data exchange result.

2. The method according to claim 1, characterized in that Based on the light reflection characteristics of different camera positions in the live sports event scene and the changes in the color temperature of the venue lighting, color adjustment parameters that match the color range distribution of the current sports picture are generated, and the color space of the original video data is nonlinearly converted using the color adjustment parameters to obtain target video data, including: Based on the light reflection characteristics of different camera positions in the live sports event scene, the light intensity distribution curve and reflection angle range of each camera position are extracted. At the same time, the color temperature values ​​of the venue lighting color temperature changes are collected to generate the current light reflection and color temperature parameter group for each camera position; Extracting a distribution histogram of the red, green, and blue color components of the current motion picture from the original video data, and calculating the adaptation weights of the red, green, and blue color components by combining the reflection angle range and the color temperature value in the light reflection and color temperature parameter group; Scaling the distribution histograms of the red, green, and blue color components based on the adaptation weights to generate color adjustment parameters that match the color range distribution of the current motion picture, the color adjustment parameters including nonlinear offset coefficients and gain coefficients of the red, green, and blue color components; The red, green and blue color components of the original video data are subjected to piecewise nonlinear mapping through the nonlinear offset coefficient and gain coefficient in the color adjustment parameters, and the original color space is converted into the target color space to generate target video data.

3. The method according to claim 2, characterized in that The method includes performing piecewise nonlinear mapping on the red, green, and blue color components of the original video data by using the nonlinear offset coefficient and the gain coefficient in the color adjustment parameters, converting the original color space into the target color space, and generating the target video data, including: Dividing the value range of each color component into a low value interval, a medium value interval, and a high value interval according to the distribution histogram value range of the red, green, and blue color components in the original video data, with the division boundaries aligned with the peak and valley values ​​of the histogram; Binding the nonlinear offset coefficient and the gain coefficient to the low value interval, the middle value interval, and the high value interval of each color component, respectively, and setting a segmented mapping rule in which the low value interval is bound to the high gain coefficient and the positive offset coefficient, and the high value interval is bound to the low gain coefficient and the negative offset coefficient; For the values ​​of the red, green, and blue color components of each pixel in the original video data, calling the bound segment mapping rule according to the value interval to which the value belongs, and mapping the target color component value; The mapped target color component values ​​are superimposed according to the scaling rule of the distribution histogram to generate superimposed values, and the superimposed values ​​are used as the color value of each pixel in the target video data to generate the target video data.

4. The method according to claim 1, wherein Adjusting the priority of coding block division based on the instantaneous motion speed and image texture complexity of moving objects in live sports events, and adaptively compensating coding quality parameters based on the brightness distribution of bright areas in the target video data, including: Extract the instantaneous speed of moving objects from live sports footage, calculate the displacement difference between adjacent video frames, and simultaneously extract the edge sharpness change of the image texture complexity to generate a joint evaluation parameter for motion speed and texture complexity. Adjusting the priority of coding block division according to the product relationship between the displacement difference and the edge sharpness change in the joint evaluation parameter to generate a coding block priority distribution map for high motion areas and high texture areas; Extracting the brightness distribution of bright areas from the target video data, calculating the area ratio of areas where the brightness value exceeds a preset highlight threshold, and generating a brightness distribution weight coefficient; According to the priority levels of different areas in the coding block priority distribution map and in combination with the brightness distribution weight coefficient, weighted compensation is performed on the compression strength and detail retention ratio in the coding quality parameters to generate adaptively compensated coding quality parameters.

5. The method according to claim 1, wherein Based on the target video data and the two-way data interaction result, a live sports event picture is constructed, and the live sports event picture is associated with the audio time series data, motion trajectory data, and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence, including: Extracting a video frame sequence after color space conversion from the target video data, and extracting a color component weight ratio, a bit stream allocation ratio parameter, and an encoding quality compensation parameter from a two-way data interaction result, and combining them with the video frame sequence to generate a live picture construction parameter group; According to the color component weight ratio in the live picture construction parameter group, the numerical distribution of the red, green, and blue color components in the video frame is adjusted, and the bitstream resources of each frame are dynamically allocated in combination with the bitstream allocation ratio parameter to generate a live sports event picture including coding quality compensation; Align each time node of the audio time series data in the live stream with the video frame timestamp of the live sports event screen frame by frame to generate synchronization timestamp parameters for the audio and video frames; According to the position coordinates of the athletes in the motion trajectory data in the live stream, the corresponding view angles in the multi-camera position space coordinates are matched to obtain the spatial anchor point parameters of the motion trajectory and camera view angle in each video frame; The sports event live broadcast images, synchronization timestamp parameters and spatial anchor point parameters are data-associated and packaged according to a preset transmission format to generate a live content transmission sequence.

6. The method according to claim 5, characterized in that Based on the athlete's position coordinates in the motion trajectory data in the live stream, the corresponding view angles in the multi-camera position space coordinates are matched to obtain the spatial anchor point parameters of the motion trajectory and camera view angle in each video frame, including: Extracting the position coordinates of the athlete corresponding to the timestamp of each video frame from the motion trajectory data in the live stream, and generating a sequence of the athlete's position coordinates based on the position coordinates; Extracting the spatial coordinates of each camera and the coverage of its shooting angle of view from the spatial coordinates of multiple camera positions; Performing position matching on each coordinate point in the player position coordinate sequence with the coverage range, and if the coordinate point is within the coverage range, marking the camera as a matching camera, thereby obtaining a preliminary matching result; Generating position mapping parameters of the relative pixel coordinates of the athlete's position in the camera image according to the relative position of each coordinate point in the preliminary matching result and the spatial coordinates of the matching camera; The position mapping parameters are bound to the spatial coordinates of the matching camera to obtain spatial anchor point parameters.

7. A system for intelligently sensing and associating multimedia content of a live sports event, configured to execute the method for intelligently sensing and associating multimedia content of a live sports event according to any one of claims 1 to 6, characterized in that: include: An acquisition module, used to obtain original video data; A conversion module is configured to generate color adjustment parameters that match the color range distribution of the current sports image based on the light reflection characteristics of different camera positions and the color temperature changes of the venue lighting in the live sports event scene, and convert the color space of the original video data using the color adjustment parameters to obtain target video data; An adjustment module is used to adjust the priority of coding block division according to the instantaneous motion speed of the moving objects and the complexity of the picture texture in the live sports event picture, and to adaptively compensate the coding quality parameters based on the brightness distribution of the bright areas in the target video data; An interaction module, configured to perform bidirectional data interaction based on the color adjustment parameters and the compensated encoding quality parameters, so that the color component weight ratio in the color adjustment parameters affects the code stream distribution ratio of the color components during video encoding; A construction module is used to construct a live sports event screen based on the target video data and the two-way data interaction results, and to associate the live sports event screen with the audio time series data, motion trajectory data and multi-camera position spatial coordinates in the live stream to form a live content transmission sequence.

8. A computing device, characterized in that The method comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for intelligent perception and association of multimedia content of a live sports event as described in any one of claims 1 to 6.

9. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for intelligently perceiving and associating multimedia content of a live sports event as described in any one of claims 1 to 6 is implemented.

Images