关键词: 景象匹配, 非下采样轮廓波, 相位一致性, Krawtchouk不变矩, 遗传算法

Abstract: Aiming at scene matching problem for taking infrared image as the actual data and the visible image as the referenced data, a multiresolution matching algorithm was proposed based on non-subsampled contourlet transform (NSCT). By using the transform of phase congruency transform, the difference of grayscale and contrast between infrared image and visible light image was weakened. Subsequently, the two types of images were separately transformed into non-subsampled contourlet domain and the proposed method took the Krawtchouk invariant moment as matching feature. The presented method, which used the improved genetic algorithm (GA) as searching strategy which conquered the precocious phenomenon, realized the multiresolution matching between infrared image and visible light image. The presented method used the relative coefficient of Krawtchouk invariant moment between the two types of images as fitness criterion for searching. Experimental results show that the proposed method has not only high matching accuracy and fast matching speed, but also better robustness in comparison with some classic matching algorithms, which can resist the geometric distortion of rotation for actual image.

Key words: scene matching, phase congruency, non-subsampled contourlet transform, Krawtchouk invariant moment, genetic algorithm