平面单足足迹图像识别  

Methods for plan single-footprint recognition

作　　者：苏美亚[1] 方欣欣[1] 暴田叶[1] 

机构地区：[1]内蒙古工业大学信息工程学院,呼和浩特010051

出　　处：《电子测量技术》2008年第12期96-97,118,共3页Electronic Measurement Technology

摘　　要：采用数字图像处理的PCA方法唯一地提取了单足足迹的足长和足宽。针对国内对足长足宽定义不统一、不唯一的情况,文中对足长足宽进行了再定义。单足足迹的足长定义为足迹区域中心线上的最远点和最近点之差。足宽定义为与足迹区域中心线垂直方向的最远点与最近点之差。采用最小距离分类器,以足长,足宽为描绘子进行了粗匹配的实验。实验结果表明对足长足宽再定义是可行的,足长和足宽的提取方法是可行的,提取结果是唯一的,粗匹配的准确率很高。The foot-length and foot-breadth of single-footprint are extracted uniquely by making use of Principal Components Analysis （PCA） of Digital Image Processing in this paper. To cope with the disunion and not unique definition of foot-length and foot-breadth, foot-length and foot-breadth are redefined. The foot-length is redefined as the longest distance of center line of footprint region, and the foot-breadth as the longest distance vertical in center line of footprint region. The experiment of coarse matching are carried using minimum distance classification in which foot-length and foot-breadth are descriptors. Experimental results indicate that the definition of footqength and foot-breadth is feasible, and the higher matching rate is acquired.

关 键 词：主分量分析(PCA) 单足足迹 粗匹配 足长 足宽 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]