Image processing system and method based on FPGA

Abstract

The invention relates to the technical field of image communication, in particular to an image processing system and method based on an FPGA, and the system comprises a data stream synchronization module which obtains a video frame and a timestamp through the FPGA; in the invention, the data stream synchronization ensures the matching of the video frame and the timestamp, the stability of the video stream is optimized, the image error caused by time dislocation is reduced, and the accuracy of video analysis is improved. And the transmission priority of the data packet is dynamically set, and the transmission strategy is adjusted according to the importance of the content, so that bandwidth resources are effectively utilized, the priority transmission of key information is ensured, and the overall efficiency and response speed of network transmission are improved. According to the adaptive coding, the coding bit rate and frame rate are dynamically adjusted according to the network condition, the adaptability of the data stream to different network conditions is optimized, and the balance between the video quality and the transmission efficiency is kept. Through the enhancement of image brightness, contrast adjustment and noise suppression, the image quality in a low-light environment is improved, and the availability and viewing experience of the image are improved.

Description

Technical Field

[0001] The present invention relates to the field of image communication technology, and in particular to an image processing system and method based on FPGA. Background Art

[0002] The image processing system based on FPGA (field programmable gate array) is an efficient image processing platform that uses the high programmability and parallel processing capabilities of FPGA to achieve rapid processing of image data. It is usually used in applications that require real-time image analysis, video surveillance, machine vision, etc., because FPGA can be programmed to execute complex image processing algorithms, such as image enhancement, filtering, feature extraction, etc.

[0003] However, in the existing technology, when processing videos in highly variable environments, the fixed transmission strategy or optimized encoding settings may lead to video quality degradation or transmission delay. In addition, in low-light and night-time image processing, due to the lack of effective noise suppression and dynamic range adjustment, it is difficult to provide clear image output, affecting the timeliness and accuracy of decision-making in key applications such as security monitoring. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an image processing system and method based on FPGA.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: an image processing system based on FPGA includes:

[0006] The data stream synchronization module obtains video frames and timestamps through FPGA, adjusts the timestamps to match the video stream, reorders and optimizes the frames, and obtains synchronized video stream output;

[0007] A priority encoding module receives the synchronous output of the video stream, analyzes key elements of the video content, dynamically sets the transmission priority of the data packet, obtains a priority adjustment state, and optimizes the data packet transmission path based on the priority adjustment state to obtain optimized video data;

[0008] An adaptive encoding module collects network status feedback, receives the optimized video data, adjusts the video encoding bit rate and frame rate, obtains network adaptive parameters, and adjusts the encoding output based on the network adaptive parameters to obtain adaptive video output;

[0009] The dynamic image enhancement module receives video input of nighttime environment, improves image brightness and contrast through FPGA, performs noise suppression and dynamic range adjustment, and obtains enhanced image results.

[0010] Preferably, the steps of obtaining the synchronous output of the video stream are:

[0011] Obtain video frames and timestamps, proofread each video frame and timestamp, check whether the time sequence is consistent with the recording sequence of the video frames, and obtain time-proofed video data;

[0012] Based on the time-corrected video data, the frame reordering index is calculated to determine the correct playback order of the video frames. The calculation formula is:

[0013]

[0014] Among them, R i is the reordering index of the i-th frame, τ i is the timestamp of the i-th frame, n is the total number of video frames,

[0015] is the sum of timestamps;

[0016] Based on the reordering index, frame reordering and video stream optimization are performed to obtain a synchronous output of the video stream.

[0017] Preferably, the step of obtaining the priority adjustment status is:

[0018] Receiving the video stream synchronous output, identifying key elements, the key elements including moving objects, color changes and scene switching, and obtaining key element analysis results;

[0019] Based on the key element analysis results, the transmission priority of each data packet is calculated using the following formula:

[0020]

[0021] Among them, H i is the priority of the ith data packet, Q k Score the importance of the kth key element, where m is the total number of key elements;

[0022] Based on the priority of each data packet, the transmission priority of the data packet is dynamically adjusted to obtain a priority adjustment state.

[0023] Preferably, the step of obtaining the optimized video data is:

[0024] Based on the priority adjustment state, analyzing the current network conditions and data packet transmission efficiency, determining the bottleneck or delay problem existing in the data packet transmission path, and obtaining a network efficiency analysis result;

[0025] According to the network efficiency analysis result, a transmission path is selected by identifying the bandwidth utilization and delay of each transmission path, and a path optimization decision is generated;

[0026] The path optimization decision is applied to adjust the transmission path of the data packet, reconfigure the data stream, and obtain optimized video data.

[0027] Preferably, the steps of acquiring the network adaptive parameters are:

[0028] Collect network status feedback, including current network bandwidth, latency, and packet loss rate, to obtain network performance indicators;

[0029] The optimized video data is received, and based on the network performance index, the video encoding bit rate and frame rate are calculated, and the calculation formula is:

[0030]

[0031] Among them, BR new is the adjusted video encoding bit rate, BR opt To optimize the original encoding bit rate of video data, B avail is the current available bandwidth, B peak is the network peak bandwidth, P loss is the network packet loss rate;

[0032] A network adaptive parameter is formed according to the adjusted video encoding bit rate.

[0033] Preferably, the steps of obtaining the adaptive video output are:

[0034] modifying the bit rate and frame rate settings of the encoder according to the network adaptation parameters to generate an encoder configuration update;

[0035] The encoder configuration update is applied to re-encode the video stream to be transmitted to obtain an adaptive video output.

[0036] Preferably, the steps of obtaining the enhanced image result are:

[0037] Receive video input of nighttime environment, identify dark and bright area information in the video, and obtain image brightness analysis results;

[0038] Based on the image brightness analysis result, the image brightness and contrast are adjusted by FPGA, and the adjustment calculation formula is:

[0039]

[0040] Among them, L new (i, j) is the new brightness value of the pixel at position (i, j), L(i, j) is the brightness value of the original pixel, γ is the correction coefficient used to adjust the contrast, L avg is the average brightness value of the image;

[0041] The noise in the video is removed, the dynamic range of the video pixels is adjusted, and an enhanced image result is obtained.

[0042] The present invention provides an image processing method based on FPGA, comprising the following steps:

[0043] Receive input video frames and timestamps, adjust timestamps to match video streams, perform frame reordering and output optimization, and generate video stream synchronization output; based on video stream synchronization output, analyze key elements of video content, dynamically set data packet transmission priority, adjust data packet transmission path, and generate priority adjustment status;

[0044] Collect network status feedback, receive video data in the priority adjustment state, adjust the video encoding bit rate and frame rate, adjust the encoding output based on the network status feedback, and generate network adaptive parameters;

[0045] It receives video input of nighttime environment, performs image brightness and contrast adjustment, performs noise suppression and dynamic range adjustment, and generates enhanced image results.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are:

[0047] In the present invention, data stream synchronization ensures the matching of video frames and timestamps, optimizes the stability of video streams, reduces image errors caused by time dislocation, and improves the accuracy of video analysis. The transmission priority of data packets is dynamically set, and the transmission strategy is adjusted according to the importance of the content, so as to effectively utilize bandwidth resources, ensure the priority transmission of key information, and improve the overall efficiency and response speed of network transmission. Adaptive coding dynamically adjusts the encoding bit rate and frame rate according to network conditions, optimizes the adaptability of data streams to different network conditions, and maintains a balance between video quality and transmission efficiency. Image brightness, contrast adjustment and noise suppression enhancement improve image quality in low-light environments, and improve image availability and viewing experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a module flow chart of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0050] See also Figure 1 The present invention provides a technical solution: an image processing system based on FPGA includes:

[0051] The data stream synchronization module obtains video frames and timestamps through FPGA, adjusts the timestamps to match the video stream, reorders and optimizes the frames, and obtains synchronized video stream output;

[0052] A priority encoding module receives synchronous output of a video stream, analyzes key elements of the video content, dynamically sets the transmission priority of a data packet, obtains a priority adjustment state, optimizes the data packet transmission path based on the priority adjustment state, and obtains optimized video data;

[0053] The adaptive encoding module collects network status feedback, receives optimized video data, adjusts the video encoding bit rate and frame rate, obtains network adaptive parameters, and adjusts the encoding output based on the network adaptive parameters to obtain adaptive video output;

[0054] The dynamic image enhancement module receives video input of nighttime environment, improves image brightness and contrast through FPGA, performs noise suppression and dynamic range adjustment, and obtains enhanced image results.

[0055] The steps to obtain synchronous video stream output are:

[0056] The video frame and time stamp are obtained through FPGA, and each video frame and time stamp are proofread to check whether the time sequence is consistent with the recording sequence of the video frame, so as to obtain the time-proofed video data;

[0057] Based on the time-corrected video data, the frame reordering index is calculated to determine the correct playback order of the video frames. The calculation formula is:

[0058]

[0059] Among them, R i is the reordering index of the i-th frame, τ i is the timestamp of the i-th frame, n is the total number of video frames,

[0060] is the sum of timestamps;

[0061] Based on the reordering index, frame reordering and video stream optimization are performed to obtain synchronous video stream output.

[0062] Specifically, the execution process of obtaining video frames and their timestamps first involves the detailed capture of each video frame and its corresponding timestamp, and a precise alignment check is performed between each frame and the timestamp to ensure the accuracy of the timestamp and the consistency of the video frame. Each frame of the video data must be completely matched with its timestamp. The equipment used in this process must have a high-precision time recording function to reduce time errors and ensure the consistency and reliability of data for subsequent processing. The result of time correction is to ensure that each frame of data in the video stream can correctly reflect the actual time point of its recording, which is crucial for subsequent frame reordering and synchronization processing.

[0063] formula The benefit of τ is that by calculating the square of the deviation between each frame’s timestamp and the average of the overall timestamps, the degree of deviation between each frame and the expected playback order is effectively evaluated, which helps to optimize the playback order of the video stream in practical applications and reduce playback problems caused by timestamp errors. i is obtained by recording the timestamp of the video frame data; parameter n is obtained by counting the total number of frames in the video; parameter is obtained by summing the timestamps of all video frames; the parameters is obtained by calculating the average value of all timestamps.

[0064] Calculation process: There is a video clip containing 5 frames, and the corresponding timestamps are τ = [1, 3, 2, 5, 4] seconds. Calculate the average timestamp:

[0065]

[0066] The frame reordering index is:

[0067] R1=(3-1) 2 =4

[0068] R2=(3-3) 2 =0

[0069] R3=(3-2) 2 =1

[0070] R4=(3-5) 2 =4

[0071] R5=(3-4) 2 =1

[0072] The results show that the timestamps of the second and third frames have the smallest deviation from the average value, and the playback order should be given priority. This result shows that in this way, the playback priority of each frame can be clarified, the video playback order can be optimized, and the coherence and fluency of video playback can be improved.

[0073] Based on the calculated reordering index, frame reordering and video stream optimization are performed. This process includes dynamically re-arranging video frames to ensure that they are arranged according to the priority indicated by the reordering index. The optimization of the video stream includes adjusting the playback order of frames to reduce the visual interference caused by frame disorder, thereby improving the quality of the final output video. The resulting synchronous output of the video stream ensures the smoothness and consistency of video playback, providing users with a more stable and smooth visual experience.

[0074] The steps to obtain the priority adjustment status are:

[0075] Receive video stream synchronous output, identify key elements, including moving objects, color changes and scene switching, and obtain key element analysis results;

[0076] Based on the key element analysis results, the transmission priority of each data packet is calculated using the following formula:

[0077]

[0078] Among them, H i is the priority of the ith data packet, Q k Score the importance of the kth key element, where m is the total number of key elements;

[0079] Based on the priority of each data packet, the transmission priority of the data packet is dynamically adjusted to obtain a priority adjustment state.

[0080] Specifically, the video stream is received and output synchronously. The analysis pays special attention to the identification of moving objects, the detection of color changes and the recording of scene switching. By conducting a detailed inspection of the picture content of each frame of video, it is possible to distinguish which elements change frequently, which colors change significantly within a specific time, and how the scene changes. This process does not rely on specific software or algorithms, but is identified by setting specific parameters. For example, the pixel change threshold is set to more than 1000 pixels per second, color changes are determined by the RGB value change amplitude exceeding 20%, and scene transitions are recorded by background changes of more than 50% in consecutive frames. Each identified key element will be assigned a score based on the frequency and amplitude of its changes, and recorded for the next step of processing. The compilation of these data forms the key element analysis results.

[0081] formula The benefit of Q is that it integrates the influence of multiple key video elements by weighted averaging, so that the transmission priority of the data packet more accurately reflects the importance of the video content, which helps to prioritize the key video data that affects the user experience in network transmission, thereby improving the overall quality of the video stream; parameter Q kThe importance score of the kth key element is obtained by analyzing the influence of the key elements identified in the video content (such as moving objects, color changes, scene switching, etc.). The score of each element reflects its visual and emotional influence in the video. The parameter m is the total number of key elements identified in the video. This number is determined by analyzing the video content and counting all elements marked as key by the algorithm or manually.

[0082] Calculation process: m = 5 key elements are identified in a video analysis result, and the importance scores of each element are Q1 = 3.5, Q2 = 4.5, Q3 = 2.0, Q4 = 3.0, Q5 = 5.0. Substitute them into the formula to calculate the priority H of each data packet i , the calculation process is:

[0083]

[0084] The result shows that the priority of the data packet is 36, which means that the data packet contains high-priority critical video content and needs to be processed first during network transmission.

[0085] Using the packet priority obtained in the previous step, the packet transmission order is further adjusted. The specific implementation is to create a priority queue and sort the packets from high to low priority. For each packet, the bandwidth and routing path are dynamically allocated during the transmission process according to its priority. High-priority packets will be sent through the fastest available path when the network is congested, while low-priority packets may be delayed in transmission on low-bandwidth paths. This process ensures the fast and priority transmission of key content in the video stream. The final priority adjustment state ensures continuous playback and high-quality output of the video.

[0086] The steps for obtaining optimized video data are:

[0087] Based on the priority adjustment status, analyze the current network conditions and data packet transmission efficiency, determine the bottleneck or delay problem in the data packet transmission path, and obtain the network efficiency analysis result;

[0088] Based on the network efficiency analysis results, the transmission path is selected by identifying the bandwidth utilization and delay of each transmission path, and the path optimization decision is generated;

[0089] Apply path optimization decisions, adjust the transmission path of data packets, reconfigure data flows, and obtain optimized video data.

[0090] Specifically, starting from the priority adjustment status, the real-time data packet transmission records and network node response time of the current network are collected and analyzed, the traffic and delay of each node are measured, the transmission time of data packets on different paths is compared, and nodes or links with low transmission efficiency are identified. This data directly reflects the real-time status of network operations. Through this network behavior observation, the main bottleneck location is found. This process generates detailed network efficiency analysis results, including which nodes or links cause increased delays, thereby providing a basis for subsequent optimization.

[0091] Based on the detailed network efficiency analysis results, the network path optimization design is carried out. This step involves specifically evaluating the bandwidth occupancy and delay time of each path, and selecting the path with the best performance as the preferred transmission path by numerically comparing the performance indicators of each path. This operation does not rely on any specific software or vague technical methods, but is directly calculated and compared through simple bandwidth and delay data. These decisions are based on real-time data to ensure that the selected path can provide good data transmission efficiency under the current network conditions. Ultimately, a clear path optimization decision is generated, and the preferred path and alternative path information are recorded in detail.

[0092] Based on the determined path optimization decision, the actual transmission path of the data packet is adjusted directly in the network management system. The operation includes manually configuring the routing tables of routers and switches, using specific network commands and configuration changes to ensure that the data stream is transmitted along the new optimized path. Through these specific network device operations, the adjustment effect is monitored in real time to verify whether the speed and stability of data transmission have been improved. The execution of this step is entirely based on the configuration capabilities of the network hardware and real-time network monitoring data, thereby ensuring the actual effect of the optimization measures. The final output is video data transmitted through the optimized path.

[0093] The steps for obtaining network adaptive parameters are:

[0094] Collect network status feedback, including current network bandwidth, delay, and packet loss rate, to obtain network performance indicators;

[0095] Receive optimized video data and calculate the video encoding bit rate and frame rate based on network performance indicators. The calculation formula is:

[0096]

[0097] Among them, BR new is the adjusted video encoding bit rate, BR opt To optimize the original encoding bit rate of video data, B avail is the current available bandwidth, B peak is the network peak bandwidth, P loss is the network packet loss rate;

[0098] A network adaptive parameter is formed according to the adjusted video encoding bit rate.

[0099] Specifically, network status feedback is collected, including current network bandwidth, delay and packet loss rate. These feedback data are obtained through a real-time monitoring system. The real-time performance data of the network is recorded regularly and analyzed to obtain a real-time network status report. The report records various network performance indicators in detail, including but not limited to the network's bandwidth utilization, delay time and packet loss rate. These performance indicators provide a reliable basis for optimizing network transmission and adjusting transmission strategies, thereby obtaining network performance indicators.

[0100] formula The benefit is that it takes into account the real-time available bandwidth of the network, the peak bandwidth of the network, and the packet loss rate of the network. It can more flexibly adjust the video encoding bit rate to adapt to different network conditions and ensure the efficiency and quality of video transmission.

[0101] Calculation process: The available bandwidth B of the current network avail The network peak bandwidth is 50Mbps. peak The data rate is 100Mbps, and the packet loss rate is P loss 0.02, optimize the original encoding bit rate BR of the video data opt is 8Mbps, then:

[0102]

[0103] The result shows that under the given network conditions, the new video encoding bit rate should be adjusted to 3.579Mbps, which means that in order to adapt to the current network conditions, the video encoding bit rate is greatly reduced compared to the original 8Mbps to reduce video playback problems caused by insufficient bandwidth or high packet loss rate.

[0104] According to the adjusted bit rate and the corresponding frame rate adjustment, a set of network adaptive parameters is formed. This set of parameters will be directly used for subsequent video encoding to ensure that the quality of the video stream matches the network capacity. Through the adjustment of these parameters, the bit rate and frame rate of video encoding can be dynamically optimized, so as to provide a better video viewing experience under different network conditions, and obtain network adaptive parameters.

[0105] The steps to obtain adaptive video output are:

[0106] Modify the encoder bit rate and frame rate settings based on the network adaptation parameters and generate encoder configuration updates;

[0107] Apply the encoder configuration update and re-encode the video stream to be transmitted to obtain adaptive video output.

[0108] Specifically, based on network adaptive parameters, such as the adjusted video encoding bit rate and frame rate, which are dynamically adjusted by the preceding module based on real-time network conditions such as bandwidth and latency, the most suitable video transmission settings under the current network conditions are evaluated to determine whether the adjusted bit rate and frame rate meet the network load capacity. For each parameter, the optimal value that adapts to the current network state is calculated. This step involves comparing the relationship between the existing bandwidth and the video data stream demand, and adjusting the encoding settings to match this relationship. In this way, the encoder configuration is updated so that each adjustment accurately reflects the change in the network state, generating an encoder configuration update;

[0109] Apply the updated encoder configuration, reset the video encoder, and apply the updated bit rate and frame rate to the actual encoding process of the video stream. This process involves compressing and processing the video data through the new encoding parameters to ensure that each frame of data complies with the new network adaptation parameters. The adjusted encoding output is designed to optimize the overall quality and transmission efficiency of the video stream. Through this re-encoding, the video data is adjusted to adapt to the changing network conditions, thereby generating the final adaptive video output to ensure the continuity and clarity of the video transmission process.

[0110] The steps to obtain the enhanced image results are:

[0111] Receive video input of nighttime environment, identify dark and bright area information in the video, and obtain image brightness analysis results;

[0112] Based on the image brightness analysis results, the image brightness and contrast are adjusted through FPGA. The adjustment calculation formula is:

[0113]

[0114] Among them, L new (i, j) is the new brightness value of the pixel at position (i, j), L(i, j) is the brightness value of the original pixel, γ is the correction coefficient used to adjust the contrast, L avg is the average brightness value of the image;

[0115] Remove noise from the video, adjust the dynamic range of video pixels, and obtain enhanced image results.

[0116] Specifically, when receiving video input in a low-light environment, the video input first undergoes a preliminary image brightness and contrast assessment, which includes automatic analysis using image processing software. The software identifies key dark and bright area information by analyzing the histogram of the image. The identification of dark and bright areas is based on statistical data of pixel brightness distribution. The distribution statistics are obtained by aggregating and calculating pixel values ​​of the entire video frame. The obtained results provide a preliminary visual impression of the light and dark contrast in the image, and finally generate a preliminary image brightness analysis result;

[0117] formula The benefit is that by adjusting the γ coefficient and normalizing it relative to the average brightness, the contrast adjustment of the image is more flexible and adaptable to different viewing conditions, especially in an environment with uneven lighting, which can significantly improve the visual effect.

[0118] Calculation process: The brightness value of a pixel L(i,j) in the original image is 50, and the average brightness of the entire image is L avg is 75, the contrast adjustment coefficient γ is 0.8, and the calculation is:

[0119]

[0120] Calculate 50 0.8 ≈34.29;

[0121] L new (i,j)=34.29·1.33=45.60

[0122] This brightness can moderately enhance the light and dark levels of the image.

[0123] A filtering algorithm is used to remove noise from the video. The filtering algorithm is used to reduce image noise. The filtering process includes analyzing the color values ​​of the pixels around each pixel, calculating the average of the color values ​​of adjacent pixels, and comparing it with the color value of the current pixel to decide whether to modify it to reduce noise. In addition, the dynamic range is adjusted by modifying the upper and lower limits of the pixel brightness to adapt to the human eye's perception of brightness changes. The output image is visually smoother and more detailed, achieving an enhanced image result. The algorithm execution efficiency can be improved through FPGA.

[0124] The present invention provides an image processing method based on FPGA, comprising the following steps:

[0125] Receive input video frames and timestamps, adjust timestamps to match video streams, perform frame reordering and output optimization, and generate video stream synchronization output; based on video stream synchronization output, analyze key elements of video content, dynamically set data packet transmission priority, adjust data packet transmission path, and generate priority adjustment status;

[0126] Collect network status feedback, receive video data in the priority adjustment state, adjust the video encoding bit rate and frame rate, adjust the encoding output based on the network status feedback, and generate network adaptive parameters;

[0127] It receives video input of nighttime environment, performs image brightness and contrast adjustment, performs noise suppression and dynamic range adjustment, and generates enhanced image results.

Claims

1. An image processing system based on FPGA, characterized in that: The system comprises: The data stream synchronization module obtains video frames and timestamps through FPGA, adjusts the timestamps to match the video stream, reorders and optimizes the frames, and obtains synchronized video stream output; A priority encoding module receives the synchronous output of the video stream, analyzes key elements of the video content, dynamically sets the transmission priority of the data packet, obtains a priority adjustment state, and optimizes the data packet transmission path based on the priority adjustment state to obtain optimized video data; An adaptive encoding module collects network status feedback, receives the optimized video data, adjusts the video encoding bit rate and frame rate, obtains network adaptive parameters, and adjusts the encoding output based on the network adaptive parameters to obtain adaptive video output; The dynamic image enhancement module receives video input of nighttime environment, improves image brightness and contrast through FPGA, performs noise suppression and dynamic range adjustment, and obtains enhanced image results.

2. The FPGA-based image processing system according to claim 1, characterized in that: The steps for obtaining the synchronous output of the video stream are: Obtain video frames and timestamps, proofread each video frame and timestamp, check whether the time sequence is consistent with the recording sequence of the video frames, and obtain time-proofed video data; Based on the time-corrected video data, the frame reordering index is calculated to determine the correct playback order of the video frames. The calculation formula is: Among them, R i is the reordering index of the i-th frame, τ i is the timestamp of the i-th frame, n is the total number of video frames, is the sum of timestamps; Based on the reordering index, frame reordering and video stream optimization are performed to obtain a synchronous output of the video stream.

3. The FPGA-based image processing system according to claim 1, characterized in that: The steps for obtaining the priority adjustment status are: Receiving the video stream synchronous output, identifying key elements, the key elements including moving objects, color changes and scene switching, and obtaining key element analysis results; Based on the key element analysis results, the transmission priority of each data packet is calculated using the following formula: Among them, H i is the priority of the ith data packet, Q k Score the importance of the kth key element, where m is the total number of key elements; Based on the priority of each data packet, the transmission priority of the data packet is dynamically adjusted to obtain a priority adjustment state.

4. The FPGA-based image processing system according to claim 1, characterized in that: The steps for obtaining the optimized video data are as follows: Based on the priority adjustment state, analyzing the current network conditions and data packet transmission efficiency, determining the bottleneck or delay problem existing in the data packet transmission path, and obtaining a network efficiency analysis result; According to the network efficiency analysis result, a transmission path is selected by identifying the bandwidth utilization and delay of each transmission path, and a path optimization decision is generated; The path optimization decision is applied to adjust the transmission path of the data packet, reconfigure the data stream, and obtain optimized video data.

5. The FPGA-based image processing system according to claim 1, characterized in that: The steps of obtaining the network adaptive parameters are as follows: Collect network status feedback, including current network bandwidth, delay, and packet loss rate, to obtain network performance indicators; The optimized video data is received, and based on the network performance index, the video encoding bit rate and frame rate are calculated, and the calculation formula is: Among them, BR new is the adjusted video encoding bit rate, BR opt To optimize the original encoding bit rate of video data, B avail is the current available bandwidth, B peak is the network peak bandwidth, P loss is the network packet loss rate; A network adaptive parameter is formed according to the adjusted video encoding bit rate.

6. The FPGA-based image processing system according to claim 1, characterized in that: The steps of obtaining the adaptive video output are: modifying the bit rate and frame rate settings of the encoder according to the network adaptation parameters to generate an encoder configuration update; The encoder configuration update is applied to re-encode the video stream to be transmitted to obtain an adaptive video output.

7. The FPGA-based image processing system according to claim 1, characterized in that: The steps of obtaining the enhanced image result are: Receive video input of nighttime environment, identify dark and bright area information in the video, and obtain image brightness analysis results; Based on the image brightness analysis result, the image brightness and contrast are adjusted by FPGA, and the adjustment calculation formula is: Among them, L new (i, j) is the new brightness value of the pixel at position (i, j), L(i, j) is the brightness value of the original pixel, γ is the correction coefficient used to adjust the contrast, L avg is the average brightness value of the image; The noise in the video is removed, the dynamic range of the video pixels is adjusted, and an enhanced image result is obtained.

8. An image processing method based on FPGA, characterized in that: The FPGA-based image processing system according to any one of claims 1 to 7 comprises the following steps: Receive input video frames and timestamps, adjust timestamps to match video streams, perform frame reordering and output optimization, and generate video stream synchronization output; based on video stream synchronization output, analyze key elements of video content, dynamically set data packet transmission priority, adjust data packet transmission path, and generate priority adjustment status; Collect network status feedback, receive video data in the priority adjustment state, adjust the video encoding bit rate and frame rate, adjust the encoding output based on the network status feedback, and generate network adaptive parameters; It receives video input of nighttime environment, performs image brightness and contrast adjustment, performs noise suppression and dynamic range adjustment, and generates enhanced image results.

Images