为了实现夜间路面物体的准确定位,提出了一种基于阴影的夜间物体单目定位技术。该技术基于张正友的平面标定方法,并采用数字图像处理技术实现了在夜间物体的准确定位。通过图像增强、大津阈值分割、形态学处理等技术提取出阴影,并利用世界坐标系与相机坐标系的关系计算出物体离相机的距离。给出了基于阴影的夜间物体单目定位技术的理论模型、基本步骤,并进行了相应的实验研究。实验结果表明:该方法能有效地实现夜间物体的定位,其中物体定位的平均误差为0.75%。相对于通常所用的双目定位方法,该方法简单易行,运算速度更快,可以推广到汽车夜间自动驾驶、机器人夜间行走等方面。

For the precise location of an object on the road at night, a positioning technology of monocular vision based on object shadow is proposed. Using Zhang Zhengyous camera calibration method and digital image processing, the technology accurately achieves positioning of objects. Through image enhancement, Otsu threshold segmentation and morphological processing techniques, extracting shadow is realized. Then, the distance between the object and the camera is calculated by the relationship between the world coordinate system and the camera coordinate system. The theoretical model and basic steps of the technology is given, and the corresponding experiments are carried out at the same time. The experimental results show that this method can effectively achieve positioning of objects at night and the average error of object position was 0.75%. Compared with the binocular vision positioning system, this method is simpler and faster, and can be applied in the automobile driving and automatic robot, etc.