红外弱小目标的匹配方法研究与仿真

Infrared Small Object Matching Method and Simulation

研究红外图像中的目标匹配问题。针对传统的红外图像中目标匹配效果不理想,匹配不准确。当红外图像中是弱小目标时,背景灰度值与目标灰度值非常相近,灰度直方图均衡化无法实现目标与背景的完整分离,导致漏匹配率较高的问题。为提高目标的匹配精度,提出一种形态学预处理的红外目标匹配方法,通过形态学算法提取出图像中的形状信息,利用形状信息将红外图像中的目标与背景完整分离,避免了只依靠灰度图像信息造成的不能完整分离目标与背景带来的漏匹配问题。进行仿真的结果证明,采用的形态学与处理的匹配方法能够完整将目标与背景分离,成功完成目标匹配,取得了满意的结果。

Research infrared image object matching. In traditional infrared image object matching, the effect is not ideal and the matching is not accurate. When infrared images are weak, the gray values of background and object are very similar, grayscale histogram equalization cannot eompletelly separate the object from the background, leading higher rate of miss matching. In order to improve the precision of the object matching, the paper put forward a pretreatment method of infrared object matching based on morphology. Through the morphological algorithm the object shape was extracted and used to separate the object from the background. The result of simulation shows that using the method, can complete separate object from background, and the results are satisfactory.