基于FPGA+DSP的全自动灯检机图像识别系统设计

Design of Automatic Light Inspection Machine Image Recognition System Based on FPGA and DSP

针对全自动灯检机对精度、鲁棒性和实时性的要求,设计一种基于数字信号处理器(DSP)和现场可编程阵列(FPGA)数字图像处理的全自动灯检机图像识别系统。该系统采用FPGA对CMOS图像传感器采集的图像信息进行预处理,并采用高性能的DSP作为图像处理的核心部分,保证了系统性能的要求。详细介绍了系统的原理和实现方案,并在硬件系统上进行了算法验证及仿真实验。试验结果表明:该系统满足精度、实时性和适应性的设计要求。

Aiming at the requirements of accuracy,robustness and high real-time of light inspection machine,the author designed a image recognition system using digital image processing technology based on FPGA and DSP.The system used FPGA to preprocess the image information capuring from CMOS image sensor,and high-performance DSP as a core part of image processing to ensure system performance requirements.This paper illustrated the principle and implementation of symtem.Moreover,in oder to validate the system,we run some algorithm verification and simulation experiment on the hardware system,the result of which showed that this system is of perfect accuracy and real-time performance,and is able to meet the design requirements.