基于双DSP的红外图像信息处理硬件平台的实现及软件优化  

Implementation of a Hardware Platform for Infrared Image Information Processing and Optimization of Software Based on double DSPs

作　　者：李坤[1] 董文娟[1] 王卫华[1] 陈曾平[1] 

机构地区：[1]国防科技大学ATR国防科技重点实验室,湖南长沙410073

出　　处：《红外技术》2008年第9期533-536,540,共5页Infrared Technology

摘　　要：介绍了一种通过实时处理来自红外探测器的高帧频红外图像序列,对空中红外目标进行自动检测与跟踪的系统。主要讨论了此系统硬件平台的设计与实现,以及算法DSP软件的设计流程。其中基于RapidIO互连的双DSP并行处理系统设计以及DSP软件优化,是算法高效实时实现的关键。因此,本文对其进行了重点论述,并提出了相应解决方法。A system, which could automatically detect and track infrared targets through real-time processing series infrared images obtained from infrared detector, is introduced. It mainly discusses the design and implementation of the hardware of this system and the design flow of the algorithms. The design of double DSPs parallel processing based on RapidIO interconnect, and optimizing software are the key factors to ensure the system to work in real-time and high efficiency. Therefore, this paper lays emphasis on these two parts and provides good measures to solve the relevant problems.

关 键 词：红外图像 实时处理 RAPIDIO 并行处理 优化 

分 类 号：TN216[电子电信—物理电子学]