太赫兹科学与电子信息学报, 2018, 16 (2): 323, 网络出版: 2018-06-09
基于协方差流形的异常驾驶行为识别方法
Abnormal driving behavior detection based on covariance manifold
异常驾驶行为识别 协方差描述子 黎曼流形 多类 LogitBoost分类器 abnormal driving behavior recognition covariance matrices Riemannian manifolds LogitBoost
摘要
研究一种高效的异常驾驶行为正确识别分类的识别方法, 对预防由于异常驾驶行为导致的交通事故具有重要意义。提出了一种新的基于协方差流形的异常驾驶行为识别方法。首先提取图像的纹理、颜色和梯度方向特征, 以克服基于单一特征识别驾驶行为的不足; 并利用协方差流形进行多特征融合, 以消除特征冗余以及不同特征数值悬殊对图像识别的影响; 最后使用多类 LogitBoost分类器进行分类识别。针对相同检测目标的正确识别率可达 98%以上, 对不同检测目标的正确识别率可达 70%以上。实验结果表明该方法有效提高了驾驶行为识别的效果。
Abstract
Abnormal driving behavior recognition is to find a method to recognize abnormal driving behaviors correctly by analyzing the driver’s activities using image processing and pattern recognition technology. This method is composed of a structure of covariance matrices of image features, which is able to extract information from data. The proposed classification framework consists in a new multi-class boosting method, working on the manifold Sym+d of symmetric positive definite d*d (covariance) matrices. The correct recognition rate for the same target can reach 98%, and above 70% for different targets. The result shows that this method effectively improves the accuracy of abnormal driving behavior recognition.
李此君, 刘云鹏. 基于协方差流形的异常驾驶行为识别方法[J]. 太赫兹科学与电子信息学报, 2018, 16(2): 323. LI Cijun, LIU Yunpeng. Abnormal driving behavior detection based on covariance manifold[J]. Journal of terahertz science and electronic information technology, 2018, 16(2): 323.