Urban surface feature refined classification method combining airborne LiDAR point cloud data and aerial images

Abstract

The invention discloses an urban surface feature refined classification method combining point cloud data of an airborne laser radar (LiDAR) and a high-definition aerial image, and aims to realize three-dimensional refined extraction of urban complex surface features in a stepped manner by adopting effective classification features for different surface features. The method comprises the followingsteps: firstly, point cloud hole tracking is carried out to realize extraction of a water body, and an improved point cloud filtering method of a progressive encryption triangulation network is adopted to realize extraction of ground points; then vegetation extraction is realized by utilizing a vegetation index image obtained by calculating a high-resolution aerial image, a vegetation point cloudcan be obtained by projecting an extraction result onto the point cloud, and on the basis, classification of tree, shrub and grassland point clouds is sequentially realized by adopting a top-down segmentation strategy. And finally, point cloud segmentation is performed on the remaining unclassified point clouds by adopting a clustering method of three-dimensional mark connectors, and identification rules for segmentation blocks are proposed, so that single extraction of buildings, bridges and lamp pole points is realized.

Description

technical field [0001] The invention relates to the combination of airborne LiDAR point cloud data and aerial images to realize the refined extraction of urban features, and in particular to a three-dimensional object-oriented classification method for ground features. Background technique [0002] Urban land use classification is the premise of many urban research work and urban management and planning. Using remote sensing data to realize the extraction of land use type information has always been one of the most important applications of remote sensing technology. In recent years, with the development of remote sensing technology, more and more data acquisition platforms (satellite, airborne, near-ground) and sensors (active, passive) provide people with multi-temporal and multi-resolution spatial data. Fine extraction and classification of urban features. Among them, high-resolution remote sensing images can provide sub-meter-level high-definition observation data, prov...

AI Technical Summary

Problems solved by technology

Although this method can significantly improve the classification accuracy, the classification result is still two-dimensional, and there will inevitably be information loss in the process of interpolating the three-dimensional LiDAR point cloud to obtain a two-dimensional image, and the interpolation process may also bring problems to the classification. error

There are also studies that try to map the spectral and texture information of the image to the point cloud to support fusion classification, but this feature-level fusion requires precise registration of the two data, which is difficult to achieve, which makes it difficult to achieve three-dimensional fusion classification

Images