摘要
Abstract
Addressing the issue of noise interference in point cloud images collected during the intelligent and unmanned transformation of gantry cranes,an adaptive statistical filtering and fitted plane filtering point cloud denoising algorithm combining point cloud density and volume is proposed.First,based on Euclidean distance,the noise points in the point cloud are divided into far-field noise points and near-field noise points.For far-field noise points,the optimal parameters of statistical filtering are determined through adaptive search,with point cloud density and volume as the objective functions for filtering.For near-field noise points,they are removed by calculating the distance between the target point cloud and the fitted plane.Experiments on the Stanford public dataset show that the proposed algorithm has better denoising performance than the traditional radius filtering method.The processing results of point cloud data collected from the actual operation of gantry cranes further verify its excellent denoising performance.This algorithm can effectively filter out point cloud noise,improve point cloud quality,and provide a reliable point cloud preprocessing method for the intelligent transformation of gantry cranes.关键词
点云降噪/自适应统计滤波/拟合平面滤波/点云密度/点云体积/门式起重机Key words
point cloud denoising/adaptive statistical filtering/fitted plane filtering/point cloud density/point cloud volume/gantry crane分类
机械制造
