论文部分内容阅读
障碍物检测一直是机器人研究领域中的热点,针对单传感器无法准确描述机器人的环境信息,为了进一步提高机器人障碍物的检测准确性,提出一种激光传感器和视觉传感器相融合的机器人障碍物检测方法。第一阶段采用视觉传感器检测出环境中的障碍物,得到障碍物的大概位置,第二阶段采用激光传感器对障碍物进行检测,判断可疑障碍物是否是真正的障碍物。结果表明,本文方法克服了激光传感器和视觉传感器的局限性,可以准确感知环境信息,并利用环境信息检测到机器路径上的障碍物,可以满足机器人实时避障要求。
In order to further improve the detection accuracy of robot obstacle, obstacle detection has always been a hot topic in robot research field. For a single sensor, it is impossible to accurately describe the robot’s environmental information. A detection method of robot obstacle based on laser sensor and visual sensor is proposed . In the first stage, the visual sensor detects the obstacle in the environment and obtains the approximate position of the obstacle. In the second stage, the laser sensor is used to detect the obstacle to determine whether the suspected obstacle is a real obstacle. The results show that the proposed method overcomes the limitations of laser sensors and visual sensors and can accurately sense the environmental information. The environmental information can be used to detect obstacles on the path of the robot to meet the real-time obstacle avoidance requirements of robots.