一种改进的块匹配运动目标识别算法  

Improved moving target recognition algorithm based on block matching

作　　者：马礼举 杨向群 

机构地区：[1]驻八四七厂军事代表室,陕西西安710043

出　　处：《应用光学》2013年第5期791-795,共5页Journal of Applied Optics

摘　　要：鉴于传统的块匹配法在图像特征提取、匹配以及搜索等方面有诸多不足,复杂的边缘提取处理与基于灰度的匹配准则都使得系统的处理速度与精度降低,为了使识别处理更加快速精确,提出了一种改进的块匹配识别算法。通过拉普拉斯算子对图像特征进行增强,提高图像的特征信息;实验表明,针对320像素×240像素大小的图像,改进算法的特征增强计算时间在Matlab中平均约为0.005 4s。Traditional block matching algorithms have many deficiencies in image feature extrac- ting and image matching, the complex edge extraction and the matching principle based on gray scale make the processing speed and accuracy decrease. To make the identification process more quickly and accurately, a recognition algorithm of the improved block matching was pro- posed. The image feature was enhanced by Laplace operator to improve the feature informa- tion. The results show that the average computing time of feature enhancement of this algo- l rithm is about O. 005 4 s for 320 × 240 pixels picture in Matlab.

关 键 词：图像处理 目标识别 块匹配法 特征增强 

分 类 号：TN911.73[电子电信—通信与信息系统] TP301.6[电子电信—信息与通信工程]