Method for identifying vegetation from airborne laser point cloud data

Abstract

The invention relates to a method for identifying vegetation from airborne laser point cloud data. The method includes the following steps that: morphological filtering is performed on LiDAR data, so that a ground object point data set can be obtained; a region growing method is utilized to segment the ground object point data set, so that a plurality of individual ground object point sets can be obtained; the spatial shape indexes of each ground object point set are obtained; and ground object point sets of which the spatial shape indexes range from 1.68 to 1.92 are judged as vegetation points and are extracted out. According to the method of the invention, three-dimensional region segmentation is adopted, and the spatial shape indexes used for describing the spatial shape features of ground objects are designed, and therefore, the method for identifying vegetation from airborne laser point cloud data can be realized according to the spatial shape indexes; means such as point cloud filtering, region growing segmentation, spatial shape index evaluation and vegetation ground object extraction is adopted, and therefore, LiDAR vegetation points can be obtained completely and accurately. The method can meet actual needs in the aspects of extraction completeness degree, accuracy and speed.

Description

technical field [0001] The invention relates to a method for identifying vegetation from airborne laser point cloud data, and belongs to the method field of LiDAR data feature extraction and processing technology. Background technique [0002] LiDAR (Light Detection And Ranging) is an emerging earth observation technology, which provides a technical means to obtain three-dimensional information of the earth's surface in a fast, high-precision and real-time manner. Unlike traditional observation technologies, LiDAR obtains three-dimensional A large number of surface feature points in space can accurately provide the spatial location information of the ground and ground objects, and it has been widely used in digital surveying and mapping, forestry monitoring, resource census and other aspects. Since LiDAR acquires the point cloud data of the entire observation area, the attribute analysis and category discrimination of LiDAR point cloud are of great significance to the applic...

AI Technical Summary

Problems solved by technology

The core idea of ​​most of the above vegetation extraction methods is actually to draw on the classification technology of remote sensing images, and then apply it to the discrete point set of LiDAR. These methods ignore the special three-dimensional spatial distribution characteristics of LiDAR itself, and Most LiDAR data often do not have the characteristics of full waveform and multiple echoes, and cannot use special properties to extract LiDAR point cloud vegetation. Therefore, the existing methods for LiDAR vegetation acquisition have many defects and cannot meet the requirements of practical applications.

Images