Study And Application On The FPGA Based Video Image Detecting System

The speed of image processing frequently touches the processing performance in imaging processing and detecting system. In this paper, based on video detecting method and embedded system and image processing, on the background of video detection in traffic information, the FPGA based video image detecting system is studied, it has higher real-time processing capability through emulation, and is able to provide the moving target's information steadily. So it is a better choice for FPGA to promote real-time performance of video processing.This paper is mainly consist of the following parts:Firstly, the whole architecture of system is given based on embedded parallel data processing system. Considering with dependability and stability, FPGA chip is used for the high performance and low cost of this system.Secondly, on the way of modularization design method, the embedded video detection system hardware platform is constructed. This system is composed of asynchronous FIFO module, image space converting module, SRAM imaging frame storing control module, image pre-processing module and image detecting module. It can resolve problems such as image sampling store, processing and translate.Thirdly, some algorithms about image processing are studied, then comparing their advantage and disadvantage, it is provided that fast median filter on FPGA for image low-level processing and its hardware structure diagram, then realizes it by VHDL language. It results as this fast median filter can lower the system resources share and promote system speed, enhance real-time performance. Fourthly, the traffic vehicle flow detecting algorithm is studied, and background differential method, image binary, and its limitation definition are discussed importantly. Their hardware structures are designed on FPGA, realized by VHDL, and made functional emulation and performance analysis.Fifthly, in system simulation design, VHDL Test Bench and TEXTIO package is used to solve the system input problem successfully. Finally, the system test results indicate that the moving object can be detected accurately and can be eliminated noises effectively.This paper mainly studies image processing and detecting technology in FPGA on-chip system, realizes all algorithms by VHDL hardware description language, and put forward some improving algorithms aiming at FPGA system characteristics. The system is emulated by using the combination of MATLAB, QuatrusII and ModelSim software, the emulating result is reached anticipated effect. This system can also be used in circumstance that needs the image to be processed and detected fast.