Dual-band infrared dim target detection based on hybrid fusion algorithm

Huang Hao; Tao Huamin; Chen Shangfeng

Dim target fusion detection in dual-band IR image is one of the key technology and research hotspot of infrared homing system. In order to establish an effective fusion-detection structure, the Boolean logic rules based fusion algorithm was analyzed. According to the principle of dim infrared target imaging, a hybrid fusion detection algorithm with AND logic and local gray was proposed based on the full use of local gray feature of dim target. First of all, infrared image of each band was segmented. Secondly, the detection results of two bands were fused using AND logic. Thirdly, to improve the probability of detection, the point beyond AND logic was determined using local gray feature. Finally, in order to optimize the local detector, the segmentation threshold in the first step was adjusted based on the fusion results. Simulation results show that the algorithm can improve the dual-band infrared detection capabilities effectively. Further, the algorithm features simple architecture and high computation speed, which makes it highly practical.

1. National Key Laboratory of Automatic Target Recognition(ATR),National University of Defense Technology,Changsha 410073,China

Received Date: 2014-01-15

Rev Recd Date: 2014-02-23

Publish Date: 2014-09-25

Key words:

Abstract: Dim target fusion detection in dual-band IR image is one of the key technology and research hotspot of infrared homing system. In order to establish an effective fusion-detection structure, the Boolean logic rules based fusion algorithm was analyzed. According to the principle of dim infrared target imaging, a hybrid fusion detection algorithm with AND logic and local gray was proposed based on the full use of local gray feature of dim target. First of all, infrared image of each band was segmented. Secondly, the detection results of two bands were fused using AND logic. Thirdly, to improve the probability of detection, the point beyond AND logic was determined using local gray feature. Finally, in order to optimize the local detector, the segmentation threshold in the first step was adjusted based on the fusion results. Simulation results show that the algorithm can improve the dual-band infrared detection capabilities effectively. Further, the algorithm features simple architecture and high computation speed, which makes it highly practical.

HTML

Reference (15)

[1]
[2]	Liu Yunlong, Xue Yuli, Yuan Suzhen, et al. Infrared smalltargets detection using local mean [J]. Infrared and LaserEngineering, 2013, 42(3): 815-822. (in Chinese)
[3]
[4]	Jiang Yue, Deng Lei, Xu Shengqiu. Review of detectionalgorithm for IR small target [J]. Infrared and LaserEngineering, 2009, 38(11): 365-368. (in Chinese)
[5]	Zhang Guangming, Sheng Weidong, Fan Shiwei, et al.Analysis of detection performance of infrared image sequencebased on multi-sensor data fusion algorithm [J]. Journal ofInfrared and Millimeter Waves, 2009, 28 (1): 18-19. (inChinese)
[6]
[7]	Zhang Lu. Research on some key techniques for imagehoming guidance based on strap-down platform [D].Changsha: National University of Defense Technology, 2010.(in Chinese)
[8]
[9]	Chan D S K, Langan D A, Staver D A. Spatial processingtechniques for the detection of small targets in IR clutter[C]//SPIE, 1990, 1305: 53-62.
[10]
[11]
[12]	Li Yang. Modeling and simulation of infrared image of deepspace target [D]. Changsha: National University of DefenseTechnology, 2010. (in Chinese)
[13]	Kim S. Scale invariant small target detection by optimizingsignal-to-clutter ratio [J]. Pattern Recognition, 2012, 45:393-406.
[14]
[15]	Zhang Bing. Point target detection and recognition algorithmsin optical image terminal homing system [D]. Changsha:National University of Defense Technology, 2005. (inChinese)

Dual-band infrared dim target detection based on hybrid fusion algorithm

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views