| 一种新的参数优化相关匹配方案 |
| |
| 引用本文： | 周健 卞松龄. 一种新的参数优化相关匹配方案[J]. 红外与激光工程, 1998, 27(1): 33-37 |
| |
| 作者姓名： | 周健 卞松龄 |
| |
| 作者单位： | 南京理工大学电光学院，中国人民解放军郑州高炮学院 |
| |
| 摘 要： | 用一组可变参数表示目标的可变模板，结合目标图像的边缘信息和模板的边界信息，定义一边缘相关能量函数，利用外部惩函数法，建立相关切配问题的总目标函数，然后用改进的模拟退火（ＳＡ）算法求得最佳匹配位置。实验结果显示，这种方案具有良好的匹配效果和较大的应用价值。
|
| 关 键 词： | 可变模板 目标初定位 模拟退火算法 图像识别 |
A NOVEL CORRELATION MATCHING SCHEMEBASED ON PARAMETER OPTIMIZATION |
| |
| Zhou Jian, Bian Songling, Tao Chunkan. A NOVEL CORRELATION MATCHING SCHEMEBASED ON PARAMETER OPTIMIZATION[J]. Infrared and Laser Engineering, 1998, 27(1): 33-37 |
| |
| Authors: | Zhou Jian Bian Songling Tao Chunkan |
| |
| Abstract: | In this paper, a deformable template is characterized via a set of variable parameters . An edge correlative energy function which combines t he edge gradient information of obj ect image and the boundary orientation information of deformable template is defined . The total object function of correlation matching problem is constructed by the external penalty function method . Then, the opti - mal matching solution of problem is achieved using improved simulated annealing ( SA) algorithm . Ex- periment results show that this scheme is efficient and practical for images with complex background . |
| |
| Keywords: | Deformable template Target inital localization Edge correlative energy function Penalty function Simulated annealing algorithm |
| 本文献已被 CNKI 维普 等数据库收录！ |
