利用计算机视觉方法对驾驶员进行驾驶状态监测和预警过程中,面部特征的快速准确定位是技术关键。针对驾驶室背景的视频帧图像,提出一种基于数学形态学,面部结构特征和灰度积分投影的方法,实现了对驾驶员两眼睛中心和两鼻孔的精确定位。实验表明,该方法具有较高的准确率和鲁棒性。
In the process of supervising driver' s condition with computer vision, fast and accurate location of facial features is the key technology. In view of the video sequence with the background of a driver' s cab, a method based on mathematical morphology, facial structure and gray integral projection is proposed to achieve the precise positioning of the driver' s eyes. The experiment indicates that this method has a high rate of accuracy and robustness.