摘要
本文提出了一种新的、利用街区距离模板搜索跟踪二值图像目标质心的快速算法,对算法原理进行了详细分析.避免了求欧氏距离的乘方与开方运算.新算法的搜索跟踪速度比文献[11]的算法快3到8倍.
A new and fast algorithm is posed in this paper, which can search and track the object, centroid of a binary image by a symmetrical and weighted mask with city block distance. The principle of the algorthm is discussed in detail. Because of neither squaring nor extracting a root, the algorithm is quickened by 300-800%.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1998年第2期161-168,共8页
Pattern Recognition and Artificial Intelligence
基金
黑龙江省自然科学基金
关键词
目标识别
跟踪
质心搜索
图像处理
二值图像
Computer Vision, Object Recognition, Object Tracking, Centroid Searching
