A Method of Change Detection in Remote Sensing Images Based on Region Correlation and Visual Words

Abstract

The invention discloses a remote sensing image change detection method based on region correlation and visual words, which is specifically as follows: first, select a reference image and an image to be detected; perform histogram equalization processing on the reference image and the image to be detected; The reference image and the image to be detected are respectively subjected to morphological reconstruction watershed segmentation; the two images obtained after segmentation are fused to obtain a fused image; the fused image includes a total of k sub-regions, and for each sub-region j, calculate the weight Correlation coefficient Cj; set the weighted correlation coefficient threshold Δ, if the weighted correlation coefficient Cj of the jth subregion is greater than Δ, then the jth subregion is the invariant region, and the invariant region is directly extracted; if the jth subregion The weighted correlation coefficient Cj of is less than Δ, then it is extracted by the method based on visual word theory. Using this method can improve the real-time and robustness of detection on the basis of meeting the detection task.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, and in particular relates to a change detection method based on region correlation and visual words. Background technique [0002] Multi-temporal remote sensing image change detection is an important direction of remote sensing technology application. For remote sensing images acquired at different times in the same area, different areas in the two images are automatically extracted and marked. This technology plays an important role in both civilian and military fields, such as disaster prevention and mitigation, earthquake early warning and battlefield damage assessment. [0003] At present, a lot of research work has been carried out in the field of change detection, and a variety of change detection methods have been proposed. Common change detection methods can be divided into three categories: the first category is direct comparison method, such as image difference...

AI Technical Summary

Problems solved by technology

The image difference method is a traditional change detection method. The change area is obtained by setting the gray threshold of the difference image, but the effect is usually not ideal; the correlation coefficient method uses the characteristics of small correlation of the change area to extract change information, which has low complexity and is insufficient. It is necessary to select a reasonable calculation area; while the classification method can obtain the changed category information, but the classification accuracy directly affects the result of the change detection, and usually the supervised classifier needs accurate change samples, which are difficult to extract, which affects this real-time

The disadvantage is that modeling is difficult, and general models such as Markov models are relatively complex, and the amount of calculation is very large, which cannot meet the requirements of fast and real-time processing.

Images