RandLA-Net outdoor scene semantic segmentation method based on local feature enhancement

Abstract

The invention discloses a RandLA-Net outdoor scene semantic segmentation method based on local feature enhancement, and the method comprises the steps: firstly, extracting the spatial geometric features of a point cloud through coding the coordinates, relative positions and normal vector direction changes of a central point and an adjacent point, and carrying out the stacking of a central point feature and a neighborhood point feature, thereby obtaining an extracted semantic feature; and connecting the geometric features and the semantic features to obtain enhanced features. Secondly, learning the weight of each feature through an attention mechanism, multiplying the weights by the features, and summing the weighted features to obtain aggregated features; then, the receptive field of each point is increased by expanding a residual block, and the network segmentation precision is improved; and finally, realizing semantic segmentation of the outdoor scene by using a coding layer-decoding layer network structure. The method has a good segmentation effect on a large-scale scene, the segmentation precision is improved to a certain degree, and good robustness is achieved.

Description

technical field [0001] The invention belongs to the technical field of point cloud scene segmentation, and relates to a RandLA-Net outdoor scene semantic segmentation method based on local feature enhancement. Background technique [0002] 3D scene understanding and analysis has always been an important research content in the field of computer vision. In recent years, with the maturity of 3D data acquisition technologies such as 3D laser technology and depth cameras, 3D point cloud processing methods have also developed rapidly. Among them, the semantic segmentation of point clouds, as the core technology of point cloud data processing and analysis, is of great significance for the mining of deeper spatial information, and plays an important role in the fields of autonomous driving, intelligent robots, medical image segmentation, etc. application prospects. [0003] Point cloud scene semantic segmentation refers to dividing the scene into several regions with different se...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a RandLA-Net outdoor scene semantic segmentation method based on local feature enhancement, which solves the problem that the semantic segmentation accuracy of the RandLA-Net network is not high due to insufficient local features.

Images