An obstacle detection method based on multiframe point cloud fusion and ground plane estimation is proposed. First, the robot, three-dimensional lidar, and two-axis rotating platform motion models were established. With the pitch rotation of lidar and the movement of the robot, the multiframe point cloud data were superposed to obtain a dense point cloud. Then, the incremental linear fitting method was adopted to estimate the ground plane to eliminate the ground disturbance. Finally, the pairwise linkage clustering method could cluster multiple obstacles and the projection method could obtain the specific size information of obstacles. Therefore, verification of experiments was carried on the Robotics Lab Platform (BIT-NAZA), experimental results showed that multiframe point cloud data can be accurately integrated to obtain dense point cloud and then get information of obstacles. For some low-height obstacles with sparse density of point clouds, the detection accuracy of an object with a height of about 10 cm is about 95%, whereas for other higher objects, the detection rate will be higher. So the proposed algorithm can achieve its detection and related parameter acquisition efficiently.
You have requested a machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Neither SPIE nor the owners and publishers of the content make, and they explicitly disclaim, any express or implied representations or warranties of any kind, including, without limitation, representations and warranties as to the functionality of the translation feature or the accuracy or completeness of the translations.
Translations are not retained in our system. Your use of this feature and the translations is subject to all use restrictions contained in the Terms and Conditions of Use of the SPIE website.