Three Dimensional Point Cloud Segmentation Method for Hydropower Plant Booster Station Based on Improved Pointcnn Model
The article proposes a point cloud segmentation method for hydropower plant booster station equipment based on an improved PointCNN model.Firstly,the PointCNN model is used to process complex and unordered point clouds.Local and global features of the point cloud are extracted through X-Conv layer operations.The PointCNN model is optimized for the complex geometric structure of the booster station equipment,and its performance in processing complex device point cloud data is enhanced by introducing self attention mechanism.The experimental results show that the improved model has achieved significant results in the segmentation task of the point cloud dataset of the booster station equipment.
point cloud segmentationPointCNNself attention mechanismboosting stationdeep learning
万方数据
维普
