期刊文献+

基于B-CFAR预处理的自动匹配目标识别法

Auto-matching target recognition method based on B-CFAR pretreatment
下载PDF
导出
摘要 针对传统模板匹配识别算法中 ,受目标旋转带来的识别率与运算速度难以两全的问题 ,文章采用了一种改进算法。首先在匹配之前用双参数恒虚警检测 (B-CFAR)提取感兴趣的区域 (ROI) ,这样排除了一些干扰 ,不仅能提高识别率 ,而且对于数据量很大的合成孔径雷达 (SAR)图像匹配速度也有所提高 ,然后对图像上每点窗口内计算方向性并依此选取相应方向的模板 ,最后用相应模板对图像进行匹配。将此算法运用在复杂二维 SAR图像飞机识别中 。 An improved template matching algorithm of recognition is presented in order to achieve both a high recognition rate and a high calculation speed,which are difficult to be gotten due to object rotation by using the traditional matching algorithm. First,the method of bi-parameter constant false alarm rate (B-CFAR) detection is used to extract the region of interest(ROI) before matching, so that the recognition effect and the recognition rate are both improved for the synthetic aperture radar(SAR) images which have a great many data. Then the directions of every pixel’s window regions are calculated so as to select corresponding templates. Finally the image is matched by using the selected template. Excellent recognition effect are obtained as the improved algorithm is used to calculate the complicated two-dimensional SAR image for airplane recognition.
出处 《合肥工业大学学报(自然科学版)》 CAS CSCD 北大核心 2005年第2期164-167,共4页 Journal of Hefei University of Technology:Natural Science
关键词 合成孔径雷达图像 模板匹配 双参数恒虚警概率 目标识别 synthetic aperture radar image template matching bi-parameter constant false alarm rate target recognition
  • 引文网络
  • 相关文献

参考文献11

二级参考文献29

共引文献101

;
使用帮助 返回顶部