摘要： 车标定位是智能交通系统中一个重要的组成部分，该文针对车标在成像过程中受到光线和其他条件影响较大的情况，提出一种基于Adaboost算法的车标定位方法，将车辆彩色图像进行预处理，再使用Adaboost算法进行车标定位，该方法具有较快的定位速度和较高的准确率。实验证明，采用该方法能获得较好的车标定位效果，具有一定的实用价值。

关键词: Adaboost算法, 车标定位, 特征提取, 类Haar特征

Abstract: Vehicle-logo location is one of the most important parts of the Intelligent Transportation System(ITS). Focusing on the effect of the light on screening vehicle-logo images, the paper proposes a vehicle-logo location method based on Adaboost algorithm. After preprocessing the vehicle-logo images, Adaboost algorithm is used to locating the vehicle-logo. The method of using Adaboost algorithm in vehicle-logo location has faster location speed and higher correctness than most of the existing methods. Experimental results demonstrate that the algorithm is feasible and applicable.

Key words: Adaboost algorithm, vehicle-logo location, feature extraction, Haar-like feature

中图分类号: 

• TP18