基于取阈分解的形态学相关灰度图像识别算法改进研究  

Improvement in algorithm for gray-scale image recognition with modified morphological correlation based on threshold-decomposition concept

作　　者：余杨[1] 崔一平[2] 

机构地区：[1]南京大学光通信工程研究中心,南京210093 [2]东南大学电子工程系,南京210096

出　　处：《激光杂志》2004年第5期65-67,共3页Laser Journal

摘　　要：为改善取阈分解形态学相关 (CMC)的识别效果 ,本文从形态学角度定义了灰度联合图像相似度 ,调整了基于图像片边缘特征提取的取阈分解形态学相关算法 (RCMC) ,并提出基于二元化图像片功率谱的取阈分解形态学相关算法 (BCMC)。计算机模拟表明 。In order to improve morphological correlation based on threshold-decomposition concept(CMC)for high similarity gray-scale image recognition,gray-scale joint image similarity degree of JTC is defined based on morphological method,CMC based on image slice edge detection(RCMC)is adjusted and CMC based on binarized image slice joint power spectrum (BCMC1 and BCMC2)is proposed.Computer-simulation results indicate that the modified algorithm can improve the discrimination capabilities with respect to the high similarity gray-scale images.

关 键 词：光通信 光学模式识别 联合变换相关器 数学形态学 

分 类 号：O438[机械工程—光学工程]