摘要： 针对智能交通系统中车辆类型自动识别问题，利用车辆面积、车窗位置和车轮位置3个特征，实现车辆类型的快速分类识别。对聚类中心初始化和模糊聚类算法进行改进，提出基于粒子群优化的改进模糊C均值算法(PSO-IFCM)的识别方法，用于车牌遮挡情况下的车辆识别。实验结果表明，PSO-IFCM算法具有较好的鲁棒性。

关键词: 车牌识别, 车型识别, 聚类中心, 特征提取, 粒子群优化, 模糊C均值

Abstract: Aiming at the problem of vehicle types automatic identification of vehicle recognition system, this paper designs a more perfect vehicle type fast classification and identification method by using vehicles characteristics of area, window position and the wheel position. An identification method based on Particle Swarm Optimization-Improved Fuzzy C-means(PSO-IFCM) algorithm is presented by improving the clustering center initialization and fuzzy clustering algorithm, and is used in block license plate of the vehicle identification in Intelligent Transportation System(ITS). Experimental results indicate that PSO-IFCM algorithm has better robustness and feasibility in the traffic regulation.

Key words: license plate recognition, vehicle type recognition, clustering center, feature extraction, Particle Swarm Optimization(PSO), Fuzzy C-means(FCM)

中图分类号: 

• TP391