A knowledge-driven automatic change detection method for high spatial resolution remote sensing images

Abstract

The invention discloses an automatic change detection method of a remote-sensing image with a high spatial resolution based on multivariate feature extraction and knowledge driving, aiming at the application requirements of high spatial resolution image change detection. The automatic change detection method of the remote-sensing image with the high spatial resolution based on multivariate feature extraction and knowledge driving mainly comprises the following steps: S1, performing knowledge driving separation on land cover areas; S2, extracting features of multivariate remote-sensing images; S3, performing change detection based on multivariate features and ground object distribution knowledge; and S4, performing post-processing of change detection based on morphology and connected domain analysis and vectorizing. By means of the automatic change detection method of the remote-sensing image with the high spatial resolution based on multivariate feature extraction and knowledge driving, the false alarm rate caused by the high spatial resolution in the traditional change detection method can be reduced effectively, and the detection accuracy of the interested variational ground objects can be kept; the automatic change detection method of the remote-sensing image with the high spatial resolution based on multivariate feature extraction and knowledge driving does not need manual intervention, and the calculation speed is fast, therefore, the automatic production of massive satellite images can be met.

Description

technical field [0001] The invention belongs to the technical field of optical remote sensing image processing, in particular to a knowledge-driven automatic change detection method for high spatial resolution remote sensing images. Background technique [0002] Remote sensing earth observation technology has the advantages of short period, low cost and wide range. Especially with the rapid development of remote sensing platforms, the spatial resolution of images is gradually increasing, and the detailed information of ground features is more abundant. The use of remote sensing image change detection technology can provide data support for dynamic monitoring of land resources. [0003] Due to the improvement of spatial resolution, compared with low- and medium-resolution remote sensing images, high-spatial-resolution image change detection technology has the following difficulties: (1) Existing change detection methods are mostly based on the analysis of image spectral diffe...

AI Technical Summary

Problems solved by technology

[0003] Due to the improvement of spatial resolution, compared with low- and medium-resolution remote sensing images, high-spatial-resolution image change detection technology has the following difficulties: (1) Existing change detection methods are mostly based on the analysis of image spectral differences, but high-spatial-resolution The spectral resolution of high-resolution images is low, and the phenomenon of "same object with different spectra-same spectrum with different objects" is serious, resulting in similar spectral characteristics of different change types in the difference image, and low performance of change type discrimination; (2) The images are acquired at different times, making it difficult to distinguish Uninteresting changes caused by factors such as illumination, shooting angle, season, and interesting changes caused by human activities, the change detection accuracy is low; (3) The improvement of spatial resolution leads to a large amount of salt and pepper noise in the detection results, which affects the change detection accuracy

Images