Cultivated land change detection method and system under layered multiple decision rules

Abstract

The invention provides a cultivated land change detection method and system under the hierarchical multiple decision rules. The method can comprise the steps that: a change mask image of a current remote sensing image is extracted, grid vectorization processing is carried out on the change mask image, and a change vector pattern spot is obtained; a slow feature analysis change detection method is utilized to extract global change information of the current remote sensing image, the historical remote sensing image and the cultivated land pattern spots, and a change probability mask image is obtained, wherein the historical remote sensing image and the cultivated land pattern spots are both existing input data; and a pixel mean value of the change vector pattern spot at a corresponding position in the change probability mask image is counted, and the mean value is taken as a change probability value of the change vector pattern spot to obtain the change vector pattern spot with the change probability value. The problems of high omission ratio, low recall ratio, difficulty in small target detection and the like in the prior art are solved; the credibility of the result is improved; and meanwhile, a method support is provided for cultivated land non-agrochemical detection, land change investigation and the like.

Description

technical field [0001] The present application relates to the technical field of cultivated land change detection, in particular to a method and system for detecting cultivated land change under hierarchical multiple determination rules. Background technique [0002] In recent years, remote sensing technology has continued to improve, making the spatial resolution, time resolution and spectral resolution of the acquired images continue to increase, which can clearly reflect the detailed information of the ground objects and express the spatial structure characteristics of the target. Handling creates good conditions. However, more detailed information not only means greater processing capacity, but also makes the interval between the changed area and the unchanged area smaller, and it is more difficult to obtain the change detection result map; secondly, to obtain The data may be disturbed by light, man-made, sensors, etc., and traditional change detection methods require f...

AI Technical Summary

Problems solved by technology

However, there are problems such as high missed detection rate, low recall rate, and difficult detection of small targets. The algorithm is still in academia, and there is still a gap between the extraction results and production requirements. There is a lack of research on the detection of cultivated land changes.

Images