Research on Point Cloud Noise Reduction Method based on Multi-Frame Fusion

Aiming at the noise existing in the component point cloud model during the robot spraying process, a noise reduction method based on multi-frame point cloud fusion is proposed. First, establish a topology structure for the point cloud data; secondly, find the closest point based on the topology structure; finally, take the closest point in the point cloud as the corresponding point, and calculate the average value to complete the fusion of the point cloud and realize the noise reduction processing of the point cloud. Taking the point cloud of plate class parts and right-angle class parts as the experimental objects, the experimental results show that after nine times point cloud fusions, the accuracy of the for plate class parts is improved by about 1.8 times; for right-angle class parts, processing The large curvature part of the back point cloud does not appear over-smoothing. The algorithm achieves a good noise reduction effect and also preserves the high-frequency information of the point cloud.