摘要
在智能车视觉导航系统中,为了能够快速精确地探测到智能车前方障碍物,并恢复出障碍物表面的三维信息,提出了一种基于激光辅助的主动探测方法.该方法首先用激光发射器向空间障碍物发射多线激光光条,然后用摄像机观测图像中光条的扭曲,线条断开的位置经过一定的三角几何计算就给出障碍物的尺寸信息.实验结果表明,该方法能够快速准确地探测到障碍物并恢复出其表面三维信息,具有很强的鲁棒性.
In order to quickly and accurately detect the obstacles in front of intelligent vehicles and reconstruct the three-dimensional information of the obstacles' surfaces, a laser-assisted active detection method is put forward in this paper. First,a laser emits mutiple laser lines to the obstacles in space,and then a camera is used to observe the twisted lines in the image. The obstacle information is obtained by means of calculating the position of the broken lines with the triangle geometry relationship. The experimental results show that the method can detect obstacles and recover their surfaces' three dimensional information quickly and accurately. It has strong robustness.
出处
《西安工业大学学报》
CAS
2015年第1期39-44,共6页
Journal of Xi’an Technological University
关键词
视觉导航
障碍物探测
激光辅助
深度信息
visual navigation
obstacle detection
laser-assisted
depth information