计算机技术与发展2012，Vol.22Issue(5)：53-56,4.
基于改进K-Means算法的车辆识别方法
Method of Vehicle Detection Based on Improved K-Means Algorithm
摘要
Abstract
Vehicle detection and vehicle recognition is important for the research of road monitoring,driving safety assistance,automatic vehicle navigation and so on. In this paper, an image segmentation method based on improved clustering center, clustering number and distance measurement of classic K-Means algorithm was proposed. Then .rectangle degree,Hu moment and Affine moment of the image characteristics were extracted, and each template of different vehicle types was established, and the vehicle recognition was conducted with combination invariant moments of improved Hu moment and Affine invariant moment. Finally, experiments show that the proposed method in this paper for vehicle detection and vehicle recognition has a good reliability and robustness.关键词
K均值/马氏距离/车辆检测/Hu矩/仿射不变距/车型识别Key words
K-means/Mabalanobis distance/vehicle detection/Hu moment/ Affine invariant moment/vehicle type recognition分类
信息技术与安全科学引用本文复制引用
张明恒,王华莹,郭烈..基于改进K-Means算法的车辆识别方法[J].计算机技术与发展,2012,22(5):53-56,4.基金项目
中央高校基本科研业务费专项资金(852009) （852009）
