Laser point cloud segmentation method for adhered canopies of fruit trees

The invention discloses a laser point cloud segmentation method for fruit tree adhesion canopies. The method comprises the following steps: S1, obtaining point cloud data of a tree crown layer; s2, performing preprocessing operation (environment denoising, point cloud projection, projection gridding and middle canopy extraction) on the acquired data; s3, performing canopy segmentation: determining a boundary baseline, selecting a minimum density value in a point cloud grid layer within a fixed offset range as a canopy segmentation point, and traversing each layer to complete determination of XOZ and XOY plane projection point cloud canopy segmentation lines; and S4, performing three-dimensional point cloud segmentation by taking the two canopy segmentation lines as a reference, retaining point cloud in a segmentation intersection, and extracting point cloud data of a single fruit tree to complete segmentation. According to the laser point cloud segmentation method for the fruit tree adhesion canopies provided