Point cloud region growth optimization segmentation method combined with K-means clustering

• Main Authors: HUI ZHENYANG, LI DAJUN, WANG LEYANG, LIU BO, XIA YUANPING, LU TIEDING, LI NA
• Format: Patent
• Language: Chinese; English
• Published: 17.11.2020
• Subjects: CALCULATING; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses a point cloud region growth optimization segmentation method combined with Kmeans clustering, and the method comprises the following steps: S1, carrying out the K-means clustering of point cloud data, obtaining the centroid of each object element after clustering, carrying out the sorting of the centroids according to the elevation, and obtaining the lowest centroid point;s2, traversing all the object elements, calculating the angle and height difference between the mass center of each object element and the mass center with the lowest elevation, dividing the object element where the mass center according with the height difference threshold value and the angle threshold value is located into ground points, and dividing the object elements where other mass centersare located into ground object points; and S3, traversing the undivided ground object primitives, carrying out regional growth on the undivided ground object primitives until all the object primitivesare traversed, and ending