A method for extracting building changes in road domain based on high-resolution images

Abstract

The present invention proposes a method and device for extracting road area building changes based on high spatial resolution remote sensing images. The characteristics or key links of the method and device include: adopting an object-oriented image processing strategy, comprehensively utilizing spectral information of images and Spatial information (including structural index, spatial relationship), using a single-class classification method. In addition, in order to avoid interference (spectral similarity) on the extraction results by feature categories outside the road area, the present invention requires that existing road information be provided in advance, such as the existing road vector diagram. The range of the road domain is extracted by using the existing road vector diagram, and then only new buildings are extracted within the range of the road domain. The method and device for extracting new buildings with high spatial resolution provided by the present invention can greatly reduce the traditional method of using image visual interpretation to obtain new buildings (including illegal buildings) in the road area. Time, improve efficiency, save human resources, and can be used in road maintenance and monitoring business operation systems.

Description

technical field [0001] The invention provides a method and device for extracting road area building changes based on high spatial resolution remote sensing images, in particular to an automatic / semi-automatic extraction of newly added buildings on both sides of a road, belonging to the field of traffic remote sensing. Background technique [0002] Timely and accurate monitoring of the road area of ​​major traffic arterial roads such as national highways and provincial highways is of great significance for maintaining the road environment, road conditions and road traffic safety. According to relevant laws and regulations, construction and development are prohibited in road areas and adjacent areas. Illegal buildings seriously affect the road traffic environment and traffic safety. Some of the illegal constructions discovered so far include constructions that have not obtained construction project planning permits, or constructions that exceed standards; construction materia...

AI Technical Summary

Problems solved by technology

Most of the existing related research focuses on the use of high-resolution images to detect various disasters (such as geological disasters, floods), information extraction of road damage, road area ecological environment evaluation, road information update, etc. (Chen Shirong et al., 2008; Wu Xuesen et al., 2011), and traditional change detection methods are mostly used, mainly relying on the spectral characteristics of images and related models (Xu Yang et al., 2011; Chen Shirong et al., 2008; Ren Yuhuan et al., 2013), but did not fully consider the high The complexity of spatial resolution images and spectral features of road objects

Some foreign scholars have carried out in-depth research on the spectral characteristics of road materials, and used them in the road extraction experiment of aerial hyperspectral images (Herold et al., 2003), but they have not been applied to actual high spatial resolution images and road change detection

Images