深度图象用于机器人视觉导引

Range Image for Robot Guidance

本文讨论了激光深度图象感测器的原理.基于该原理,文中提出了一种计算时间省、抗干扰能力较强的提取图象边界的新方法.文中还提出了一种获取三维数据的变换方法和一种从表面到物体的识别方法.将上述方法和技术用于导引机器人焊接,提高了机器人适应外界环境的能力.

The principle of a range-finder is discussed and a new edge extracting method, the single side difference method is suggested. This method can be used to detect the exact position of a laser stripe in a short computing time. An algorithm for two-dimensional to three-dimensional transformation and 3-D technique for the recognition of an object from its external surface are presented. These techniques have been used in guiding a welding robot and the results are found to be fairly good.