Natural protected area human activity change detection method based on multi-scale feature fusion

Abstract

The invention provides a natural reserve human activity change detection method based on multi-scale feature fusion. The method can serve a natural reserve human activity supervision service, improves the work efficiency of the natural reserve human activity change monitoring service, and reduces the work cost. The detection method mainly comprises seven steps of data selection, data preprocessing, principal component analysis, image segmentation, spectral change vector construction, spectral change degree measurement and post-processing, and then a final change object is extracted. According to the detection method disclosed by the invention, due to the improvement of the working efficiency, human activities existing in the natural protected area can be discovered in time, and early discovery and early treatment of the human activities are realized, so that the natural protected area is better protected, and the national ecological environment safety is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of ground monitoring and ecological environment protection, and relates to a detection method for human activity changes in nature reserves based on multi-scale feature fusion. Background technique [0002] Remote sensing change detection is based on observations of the same object or phenomenon at different times to determine its different processing procedures. Remote sensing image change detection is the use of multi-source remote sensing images and related geospatial data covering the same surface area in different periods, combined with the corresponding features and remote sensing imaging mechanism, to determine and analyze the changes of the region and features, including the location and range of the features. Changes and changes in feature information and status. Its purpose is to find out the change information of interest and filter out irrelevant change information as interference factors. Remo...

AI Technical Summary

Problems solved by technology

[0009] (4) Most of them specialize in or use a certain method, and lack the methods of comprehensive integration of multiple methods and strategies, such as focusing on spectral algebra, spectral changes, or classification. Take advantage of multiple approaches

[0010] 3. In the field of remote sensing monitoring of changes in human activities in nature reserves, there are few studies on related automatic detection techniques and methods, and industry applications currently mainly rely on manual interpretation

[0011] The present invention is closely related to the urgent needs of the current field of human activity supervision in nature reserves, and aims at the lack of automatic monitoring technology for changes in human activities in natural reserves and the failure of existing remote sensing automatic change detection technologies in other fields to make full use of remote sensing. Based on the multi-scale features of images and the failure to integrate the advantages of various technical methods, etc., based on the characteristics of nature reserves and human activities (the change in human activities is relatively small, once the change is easy to form a relatively strong contrast with the surrounding background, etc.), comprehensive utilization Based on multi-scale features such as pixels, neighborhoods, objects (patch), and scenes on time-series remote sensing images, and combined with transformation methods, algebraic methods, and object-oriented multiple change monitoring methods, a multi-scale feature-based Integrated method for monitoring changes in human activities in nature reserves

Images