The Design Of Video ROI Compression Coding System Based On HI3516

With the development of the national economy and the continuous advancement of science and technology,the construction of urban safety prevention and control has been popularized on a large scale over recent years.From the nation's perspective,the Skynet and Xueliang Projects,which are deployed to meet the needs of urban security prevention and control,as well as to the schools and supermarkets around us,are inseparable from video surveillance equipment.It provides great help for police handling and traffic system violation monitoring because of the popularity of monitors with camera equipment.Currently,the encoding scheme used in video surveillance systems is a large percentage of lossy compression of the video footage,which results in the clarity of some detailed parts of the video not meeting the needs of the user.Given the lack of large-scale lossy compression of surveillance video,this article designs a video ROI(Region Of Interest)compression encoding system based on HI3516.The main control chip of the system hardware platform is Hisilicon HI3516,equipped with FLASH,DDR4 internal memory,communication serial port,Peripheral circuits such as power interface and network interface peripheral circuits.The system uses Sony's IMX-327 CMOS camera to collect video data and convert it into digital signals.The video input module of HI3516 receives the collected digital signal and converts the video to YUV format at the same time.The ROI extraction module uses the BING target detection algorithm to extract the coordinates of the region of interest in the YUV format video,and sends the coordinates to the video encoding module,compresses and encodes the video in the H.264 format,and assign a lower ROI area Quantization step size,other areas are assigned a higher quantization step size,and finally,the code stream generated after encoding is sent to the client through the RTSP network transmission protocol for decoding and playback.This coding method,which adds ROI extraction to H.264 coding can achieve high-quality,high-definition compression coding for the region of interest in the surveillance video,and low-quality compression coding for the non-interest region,ensuring clarity.Under the premise that the speed will not be affected,it occupies smaller memory.The experimental results show that this system can realize video compression coding,the clarity is not much different from H.264 compressed encoding clarity,and the compression rate is increased by about6%.This means that the method achieves its expected effect,that is,the video occupies less memory,which is conducive to the storage and network transmission of video data.The system has practical application value.