High-resolution remote sensing building extraction method combining semantic segmentation and edge detection

Abstract

The invention discloses a high-resolution remote sensing building extraction method combining semantic segmentation and edge detection, and belongs to the technical field of remote sensing image target extraction. The method comprises: firstly, by adopting a semantic segmentation network and a deep edge detection network, predicting a building roof surface and an edge graph respectively; then, denoising and skeleton extraction are carried out on the building edge graph; on the basis, in combination with a building roof surface prediction map, carrying out deburring and broken line repairing processing on the building edge map; and finally, vectorizing the building edge graph to obtain a building roof contour extraction result. According to the method, the problem that the building edge in the deep convolutional neural network model prediction map is discontinuous and not closed can be solved, the extraction precision of the building edge of the high-resolution remote sensing image is improved, and the method is good in performance and wide in applicability in extraction of roof forms of different types of buildings.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image target extraction, and in particular relates to a high-resolution remote sensing building extraction method. Background technique [0002] Building extraction based on remote sensing images provides important data support for urban and rural planning, dynamic monitoring, disaster assessment, land use analysis, map update, etc. With the continuous development of remote sensing technology, the spatial resolution of remote sensing images that people can obtain is getting higher and higher. Buildings present richer spatial and geometric features on high-spatial-resolution remote sensing images (hereinafter referred to as high-resolution remote sensing images), and the details inside the roof of buildings are also clearly visible, which brings opportunities for building extraction tasks with challenges. [0003] Traditional building extraction methods usually require manual construction ...

AI Technical Summary

Problems solved by technology

[0005] The present invention overcomes the problems of discontinuous and unclosed edges existing in the existing building extraction technology, and proposes a high-resolution remote sensing building extraction method combining semantic segmentation and edge detection

Images