摘要
分析了图像识别中模板匹配技术面临的计算量大、存储量大的问题,提出了基于遗传算法的图像识别方法。该方法首先对图像模板进行离散化处理,对图像离散点控制,从而把图像识别问题转化成一系列离散点的组合优化问题;然后利用遗传算法对种群优化的性能,对各个控制点组合优化,使各控制点与模板匹配;最后通过计算机仿真实验,证明了这种方法的有效性和实用性。
The template matching technique is confronted with large computational quantity and memory quantity in image recognition. The method of image recognition based on genetic algorithm was proposed. With this method, the image is made discrete, and the discrete point of image are controlled. As a result, the image recognition becomes of combination optimization of these discrete points. By means of the properties of genetic algorithm, combination optimization of the control points was made to match the control point with template. The results of the simulation experiment show that this method is more effective and practical than the traditional method.
出处
《长安大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2004年第6期98-101,共4页
Journal of Chang’an University(Natural Science Edition)
关键词
图像识别
模板匹配
遗传算法
组合优化
image recognition
template matching
genetic algorithm
combinational optimization
