基于NiosⅡ软核的实时红外图像自动跟踪系统设计  

Real-Time Infrared Image Automatic Track System Based on NiosⅡ

作　　者：贺明[1] 王新赛[1] 

机构地区：[1]防空兵指挥学院红外与射频技术研究中心,河南郑州450052

出　　处：《红外技术》2007年第1期34-37,共4页Infrared Technology

摘　　要：针对目前国内外大多数红外图像自动跟踪系统采用DSP+FPGA结构,系统存在设计电路复杂、成本高、实时性较差等缺点,开发出在单片现场可编程门阵列(FPGA)中实现以NiosⅡ软核处理器为核心的实时红外图像自动跟踪系统,采用自定义指令的NiosⅡ处理器实现结构复杂的自递归Otsu的聚类分割和目标运动预测算法,利用硬件实现目标的快速相关匹配,软硬件协同完成系统对目标的实时跟踪,实验结果表明,系统不但体积小、功耗低、实时性强,而且在复杂的地物背景下,能够稳定、精确地跟踪目标。According to many infrared image automatic track system using the framework of DSP＋FPGA hardware ,the systerm has the disadvantage of complex circuit, expensive price, slow speed and so on. The automatic track system based on software core Nios II is designed, it adopts custom instruction to realize real-time image recursion Otsu＇s method segmentation and target forecast, adopts hardware to realize real-time fast image matching track, the results, the system not only has small bulk, lower power, fast speed and also has better performance of image track show that on complex background.

关 键 词：FPGA NiosⅡ 相关匹配 OTSU分割 自动跟踪 

分 类 号：TN216[电子电信—物理电子学]