基于综合特征的车辆检测识别系统

随着信息技术和智能交通的迅速发展，自动的车辆检测识别成为不可或缺的技术。由于基于单一特征和单车辆的识别并不能满足实际中的应用要求，因此提出了多特征多车辆的检测识别算法，利用基于灰度对称，形状特征和光流强度特征的提取，通过基本的图像处理算法，最终由综合特征为基础得出车辆的识别检测结果，利用基于VC＋＋6．0的软件系统开发出车辆检测识别系统，通过实验对照，由单一特征和综合特征的检测不同结果获得实验数据，通过检测算法将视频序列图像中的车辆信息自动识别提取出来并加以标记。实验结果表明系统能够适应多变的车型和环境信息，通过识别结果准确性的分析，得出结论：针对不同的车型有较好的鲁棒性，能够满足实际智能交通控制平台的构建的基本要求。

Vehicle Detection and Recognition System Based on Comprehensive Features

Automatic vehicle detection and recognition has become indispensable along with the development of information technology and intelligent transportation. Because the vehicle recognition based on single feature and vehicle can not meet practical application,a new image detection and recognition method based on comprehensive features and vehicles has been proposed, using feature extraction based on grey symmetry, shape and optical flow intensity, through basic image processing algorithm, obtain the result of recognition and detec- tion based on comprehensive feature finally ;Apply VC＋＋6.0 to develop vehicle detection and recognition system. Get the experimental data according to different result between single feature and comprehensive feature, automatically recognize and extract the vehicle infor- mation in the video sequence image by detection algorithm, then lable it. The experimental results show that the system has perfect robustness to different vehicles and could meet the construction of practical intelligent transportation system.