摘要
论文在Linux平台下设计了一种目标跟踪算法。算法中使用图像的三个不变矩代替图像像素的灰度值作为特征量,利用改进后的平均绝对差算法作为相似性测度计算公式进行模板匹配;同时,为了提高目标跟踪的速度,采用遗传算法对图像的匹配策略进行优化,提高了匹配速度。与Windows平台下的目标跟踪算法相比,该算法的实时性高。
On Linux operating system, a target tracking algorithm is designed. In the algorithm, three invariant moments of image are used to take the place of gray of pixel as character, and mean absolute difference algorithm improved is used as similarity calculation formula to do template match. Otherwise, to improve the match speed, genetic algorithm is used to optimize the match strategy. Compared with target tracking algorithms on Windows operating system, the algorithm has higher real time.
出处
《微计算机信息》
北大核心
2008年第4期259-260,278,共3页
Control & Automation
基金
教育部"新世纪优秀人才支持计划"(NCET-05-0910)
关键词
目标跟踪
图像不变矩
遗传算法
target tracking
invariant moment of image
genetic algorithm
