摘要
基于数学形态滤波,提出了一种红外小目标实时检测的算法及其DSP硬件实现技术。首先分析了小目标检测过程,并论述了改进的形态滤波检测算法原理,然后根据算法特点提出了在DSP64X的实现方法和过程,最后给出并分析了该算法在红外探测系统中的运行结果。实验结果表明:该方法可以快速、可靠的检测出极低信噪比下的小目标,满足了系统性能指标。
Based on mathematic morphology filtering, a real-time detection algorithm for infrared small target and its realizing technology on DSP hardware are proposed. Firstly, the small target detecting process is analysized and the improved morphology filtering algorithm for target detection is introduced. Secondly, based on the algorithm＇g characteristic, its implementing method and procedure on DSP64X are presented. Finally, the running result of algorithm on infrared detecting system is given and analysized. The experiments show that the method can fleetly and reliably detect the small target with low SNR, which has achieved the demand of high performance.
出处
《激光与红外》
CAS
CSCD
北大核心
2006年第8期715-717,共3页
Laser & Infrared
基金
中科院创新基金项目
关键词
红外图像
形态滤波
目标检测
DSP技术
infrared image
morphology filtering
target detection
DSP technology
