Research And Implementation Of Real-Time Pedestrain Detection System Based On SoC FPGA

Cars bring great convenience to people's lives,but at the same time bring safety hazards to pedestrians on the road.In order to effectively protect pedestrians,scientists and engineers are committed to developing fast,accurate and reliable pedestrian detection systems to assist drivers in driving safely.This paper proposes a real-time pedestrian detection system based on SoC FPGA.The main research work of this paper is as follows:1.The system designed in this paper is mainly composed of image acquisition and display system part and pedestrian detection algorithm.It adopts the top-down modular design idea and divides the system into multiple sub-modules for research and design.2.This paper uses the DE10-Nano development board and D8 M camera module of Terasic Technology Company,to build a camera acquisition and display system.We will write Verilog code for the camera acquisition module,image buffer module and image display module in the system,and use Qsys software builds the SoC system.640×480 resolution video images can be stably output through on-board debugging.3.This paper uses the Histogram of Oreinted Gradient(HOG)+ Support Vector Machine(SVM)algorithm for pedestrian detection.The calculation steps of the algorithm are studied in detail.The hardware platform of FPGA is simplified and optimized appropriately.The algorithm is implemented in parallel by pipeline parallel processing.Through simulation and on-board debugging,the design of real-time pedestrian detection system is finally completed.Tested in the actual environment,the pedestrian detection system implemented in this paper can detect pedestrians more accurately,achieving a detection rate of 20 frames per second,meeting the requirements of real-time.