论文部分内容阅读
交叉熵值的大小反映了模板图像与实时图像之间的信息量差异大小,从平均意义上来表征模板图像与实时图像之间的信息量差异量。为了解决机载成像光电吊舱系统中的图像辐射失真和几何失真问题,提出了基于局部交叉熵的图像匹配跟踪算法。由于交叉熵值最小准则有利于信息量丰富的图像匹配,因此该算法不仅具有抗噪能力,而且具有良好的抗辐射失真和抗几何失真的能力。仿真试验表明:在辐射失真情况下,该算法具有稳健的匹配跟踪能力,适应能力强,是一种很实用的匹配跟踪算法。
The size of cross-entropy value reflects the difference of information amount between template image and real-time image, and represents the information amount difference between template image and real-time image in the average sense. In order to solve the problem of image radiation distortion and geometric distortion in the airborne optoelectronic pod system, an image matching and tracking algorithm based on local cross entropy is proposed. Because the minimum cross-entropy criterion is conducive to information-rich image matching, the proposed algorithm not only has the anti-noise ability, but also has good anti-radiation distortion and anti-geometric distortion ability. The simulation results show that the proposed algorithm has a good matching and tracking ability in the case of radiation distortion and has strong adaptability.