| The popularity of cars bring convenience and comfort to people’s life, butfrequent traffic accidents often make people have a lingering fear. Therefore, in thisthesis, we developed an automotive collision avoidance system for vehicle safetyissues, which can protect the driver and reduce the incidence of traffic accidents.The current embedded system and video image processing technology provides afeasible technical platform for the automotive collision avoidance system experiment.In the video image processing platform, by comparing the characteristics of the ASIC,FPGA and DSP, we chose to use FPGA, which has the characteristics of richhardware logic resources, good real-time performance and parallel processing. Interms of system control platform, we adopt the STM32which based on cortex-M3kernel. the chip has the characteristics of high performance, easy development, lowpower consumption.This system mainly includes the design of image acquisition, image processing,and the alarm system, FPGA is responsible for the core task in the process of videoimage processing, including image segmentation, filtering and edge detection.Considering the real-time system, we adopt median filtering algorithm and Robertsoperator, which is stable and easy to implement. In the design of alarm system, weuse the embedded platform based on STM32MCU. Finally， through theestablishment of automobile security model, we analyze the resulting data obtainedby image processing and make smart decisions on the embedded platform, It can helpthe driver to take appropriate hedging measures to reduce accidents.In this thesis, the collision avoidance system has the characteristics of low cost、real-time and high reliability. Through the debugging and experiments, the systemhas basically reached the expected effect and meeted the requirements of automotivecollision avoidance system. |
PDF Full Text Request