摘要
针对红外图像序列中的小目标实时检测问题,文中提出一种基于时空域融合滤波的小目标检测算法。首先介绍了时空域融合滤波检测算法的基本原理,然后根据算法的特点提出了它在多TMS320C6416上的并行流水线实现方法,最后给出并分析了算法软件仿真结果和硬件平台设计。实验结果表明该方法可以快速可靠地检测出低信噪比红外图像序列中的小目标,从而有效地验证了本系统的实时性及适用性。
For the real-time detection of small target in infrared image sequence, a temporal-spatial fusion filtering algorithm was provided. The basic theory of this algorithm is introduced. And then according to the characteristic of the algorithm a parallel pipeline method is presented based on TMS320C6416s. Finally, a hardware platform was designed after gaining and analyzing the result of simulating. The result of experiment showed that this method could detect small target in low signal-to-noise ratio infrared image sequence fast and reliably and indicated that this system had good realrime ability and applicability.
出处
《弹箭与制导学报》
CSCD
北大核心
2007年第5期75-77,共3页
Journal of Projectiles,Rockets,Missiles and Guidance