一种基于单元信息熵的目标匹配改进算法

为解决图像信息熵无法有效进行图像匹配的问题,将图像单元信息熵和投影特征相结合,定义了图像单元信息熵,并提出了一种基于单元投影信息熵的图像匹配方法.在单元信息熵的基础上,在各个单元格内进行单元信息熵投影计算,然后按照一定的测度进行计算,从而实现图像的匹配.采用网格分层的搜索算法,加快搜索速度,提高其工程实用性.实验证明:该算法具有良好的抗几何失真能力和抗辐射失真的能力,以及很好的抗噪声干扰的能力,可以准确的进行目标匹配.

Improved Target Matching Algorithm Based on the Unit Projection Entropy

In order to solve the problem that the image information entropy method worked terribly in image matching, the unit entropy and the projection features are combined to solve this problem. Then the image unit projection entropy is proposed.Based on the unit entropy, the projection vector in each unit-region is computed to realize the image matching. The gridding searching method is used to enhance the search speed,and improve the practicability. The experiment shows that the algorithm is good at radiation and anti-distortion and does well in noise capability. This algorithm can detect the target very well.