Automatic change detection method and system based on historical background and current remote sensing image

Abstract

The invention discloses an automatic change detection method and an automatic change detection system based on a historical background and a current remote sensing image. The method comprises the following steps: arranging historical background data, and carrying out the multi-scale segmentation of remote sensing image data in a new period under the guidance of the historical background data; calculating a multi-scale segmentation result; counting, analyzing and calculating to obtain multi-dimensional features, searching outliers on feature dimensions in each category, and marking the outliersas first suspected change pattern spots; and obtaining second suspected change pattern spots with different change probabilities through decision modeling according to the first suspected change pattern spots, and removing pseudo pattern spots from the second suspected change pattern spots. The method has the beneficial effects that the subjectivity of manually selecting samples in a traditionalinterpretation and analysis method is avoided, and the problem of remote sensing monitoring is solved by flexibly applying a statistical analysis method. The change detection precision is improved. Different features are selected for different ground object change detection scenes. Change pattern spots are detected through a statistical analysis method. The positive detection rate of change discovery is improved.

Description

Technical field [0001] The present invention relates to the field of remote sensing technology, in particular to an automatic change detection method and system based on historical background and current remote sensing images. Background technique [0002] The use of remote sensing technology for change detection is currently one of the hotspots in the field of remote sensing. With the continuous improvement of the temporal and spatial resolutions of remote sensing images, it is of great practical significance to combine multi-period images and auxiliary data of the same surface area to develop change detection applications. [0003] Remote sensing image change detection methods are divided into two categories: pixel-based image change detection and feature-based image change detection. Among them, pixel-based image change detection directly uses image information to obtain the change area. The advantages are simple method and fast speed. It is easy to obtain the change area. Comm...

AI Technical Summary

Problems solved by technology

[0003] The methods of remote sensing image change detection are divided into two categories: pixel-based image change detection and feature-based image change detection. Among them, pixel-based image change detection is to directly use image information to obtain the changed area. The advantage is that the method is simple and fast. It is easy to obtain the change area, and the common methods include algebraic operation method, change vector analysis method, space transformation method and classification and comparison method. "To carry out change detection for the analysis unit, it is easy to have too many false changes on the "salt and pepper" image, which requires a lot of manual interactive processing work in the later stage; multi-period image change detection is too dependent on image quality, and there is more pre-processing work in the early stage

In 2015, Yang Xiaomei proposed a land cover change detection method for high-resolution remote sensing images based on historical data mining, but the change detection was too dependent on the radiation difference between the two images, and did not analyze and process the pseudo-patterns in the later stage

Images