基于FPGA的视频图像高速处理技术——在钢轨动态检测中的应用  

Fast Processing Technology of Video Image Based on FPGA——Application in Dynamic Detection of Rail

作　　者：张欣[1] 王兵[2] 张春华[1] 

机构地区：[1]河北大学电子信息工程学院,河北保定071002 [2]河北大学数学与计算机学院,河北保定071002

出　　处：《计算机工程与应用》2006年第34期178-181,共4页Computer Engineering and Applications

基　　金：河北省教育厅基金项目资助(2003202)。

摘　　要：高速图像处理可用软件和硬件两种方案实现。软件方案成本低,灵活,但速度慢;硬件方案速度高,不够灵活且成本高。现场可编程门阵列(FPGA)正好能解决这一矛盾。介绍了一种基于FPGA的视频图像高速处理技术,它被成功地用于钢轨断面图像的实时动态监测系统中。该系统采用了投票表决算法,用VHDL语言(超高速集成电路硬件描述语言)编程,采用多语言协同仿真技术(FLI)。结果表明,该系统充分发挥了FPGA器件的并行特性,显著提高了图像处理速度,达到了动态监测的实时性要求。The fast image processing can be achieved by the software methods and the hardware methods.The software methods have the characteristics of low cost and high flexibility,but low speed.The hardware methods are on the contrary.The field programmable gate array（FPGA） can combine them in a way that conserves their respective advantages. In this paper,a fast processing technology of the video image based on FPGA is introduced,which is applied in the real-time and dynamic detection system of the rail section images successfully.In this system,the voting algorithm is adopted,the very-high-speed integrated circuit hardware description language（VHDL） is applied for programming,and the foreign language interface（FLI） is used for simulation.The results show that the parallel characteristic of FPGA is exerted adequately and the speed of the image processing is enhanced obviously in this system and the real time characteristic of the dynamic detection is achieved.

关 键 词：FPGA 视频图像 并行处理 投票算法 光取断面法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]