摘要
为了解决当前车辆目标成像效果差且光照干扰强,导致智能系统识别车辆能力不足的问题,提出了基于边缘检测与模式识别的车脸识别算法。首先,基于Lab颜色空间转换与Canny边缘检测,设计车辆前脸区域检测机制,得到车脸区域图像。然后,基于粗糙集描述和Adaboost分类器,对车脸特征完成训练,建立强识别器,准确识别车脸。最后,基于开源图像库Aforge.NET和C#语言实现算法,开发出瀑布结构标准软件系统。实验测试结果显示：与当前车脸识别技术相比,算法拥有更高的准确性与稳定性。
In order to solve the problem of poor car reorganization ability of intelligent systems induced by poor imaging performance and strong light interference, a car face recognition algorithm based on edge detection and pattern recognition is proposed in this paper. Firstly, the vehicle front area detection mechanism is designed based on Lab color space conversion and Canny edge detection to get a car face area image. Then the vehicle face features are trained based on the rough set description and the Adaboost classifier to build strong recognizer for accurately recognizing the car face. Finally, a waterfall structure standard software system has been developed based on open-source image library Aforge.NET and C# language to implement the algorithm. Experimental results show that this algorithm has higher accuracy and stability than the current mainstream vehicle recognition technoloav.
出处
《控制工程》
CSCD
北大核心
2018年第2期357-361,共5页
Control Engineering of China
