侧抑制在红外噪声图像配准中应用  

The Application of Lateral Inhibition in Infrared Noise Image Registration

作　　者：薛弘晔[1] 李言俊[1] 张科[1] 

机构地区：[1]西北工业大学航天学院,陕西西安710072

出　　处：《红外技术》2008年第4期217-220,共4页Infrared Technology

基　　金：航空科学基金(04I53067);航空支撑科技基金(05C53005)

摘　　要：图像特征配准主要使用边缘,而红外图像边缘呈渐变性,且信噪比低。首先提出了使用非循环侧抑制网络对图像进行增强处理,并改善了侧抑制模板系数矩阵,提高了红外图像信噪比。其次对使用欧氏距离进行相似度度量函数进行了改进,最后使用了蚁群算法对配准参数寻优。仿真试验表明了算法的有效性。The registration of image feature mainly uses the edge, but the edge of the infrared image shows the qualify of gradual change and the Signal to Noise ratio is lower. Firstly the non-cycle literal inhibition network to enhance the images is presented, and improve literal inhibition template coefficient matrix. It also improves the Signal to Noise ratio of the infrared image. Secondly using Euro-distance to measure the similarity of the function is refined. Finally, it uses Ant Colony Algorithm to fetch the most appropriate registering parameters. The result shows that the algorithm is available.

关 键 词：侧抑制 图像配准 蚁群算法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]