三维点云室内平面要素精确提取方法研究

打开文本图片集

中图分类号：TP399 文献标识码：A 文章编号：2096-4706（2025）24-0163-04

Abstract： Indoor model semantic reconstruction isan important basis for research fields such as smart cities，artificial intelligence，andindoornavigationandpositioning.Thispaperstudiestheindoorpointclouddata，andproposesanaccurate extraction method of indoor point cloud plane elements under geometric constraints.This method first uses the hybrid segmentation method tosegment the indoor pointcloud，andthenuses the geometriccharacteristics of the indoorsemantic components toaccuratelyextract thesegmented plane.The extractionplaneincludes ground，ceiling，wall，doorplane，beam planeand column plane.There are two groups of experimentaldata.Oneis thesingle-layerlaser pointcloud dataofateaching building，andtheotheristhe standarddataof the International SocietyforPhotogrammetryand Remote Sensing（ISPRS）.The validity and reliability of the method are verified by the evaluation of the test results.

Keywords： indoor point cloud; hybrid segmentation; plane element; accurate extraction

0 引言

随着信息技术与互联网技术的迅猛发展，智慧城市[1已迈向万物互联、万物智慧的新阶段。（剩余4548字）