Glacieridentification method based on correlation coefficient threshold

Abstract

The invention relates to a glacieridentification method based on acorrelation coefficient threshold. The method comprises steps as follows: firstly, an SAR (synthetic aperture radar) image is processed by original data reading, image registration, interferogram calculation, coherence estimation, flat ground phase removal, phase filtering and geocoding, correlation coefficientmaps are obtained, anda glacier area has a low-coherence characteristic on the correlation coefficient maps; a curvilinear relationship between coherence coefficient values and corresponding pixel number in the correlation coefficient maps in a research area is fitted, a first-order derivative of the curvilinear relationship is solved, a place where the first-order derivative curve becomes flat is taken as a transition point from the glacier area to a non-glacier area on the image, and a selected threshold b is obtained; the correlation coefficient maps are classified with a rule-based object-orientedclassification method, and the classification rule is set as follows: gamma is smaller than b; finally, the obtained image is subjected to post-classification processing including small pattern spot removal, raster-to-vector conversion, removal of abnormal holes in a planar object and serrated edge smoothing, and the glacier zone boundary is obtained.

Description

technical field [0001] The invention relates to a glacier identification method based on the coherence coefficient, in particular to the determination of the coherence coefficient threshold for identifying the glacier. [0002] technical background [0003] As an important adjustable freshwater resource, glacier resources are closely related to regional climate change and hydrological effects, and are good indicators of changes in the climate system. With global warming and shortage of fresh water resources, the monitoring of changes in glaciers, especially mountain glaciers, has become an important issue in resource utilization and sustainable development. Early glacier monitoring mainly relied on field surveys. Before the 1930s, changes in several monitoring points at the end of the glacier were relied on to calculate the change in glacier length. However, mountain glaciers are mostly distributed in areas with high terrain and steep terrain, so it is difficult to carry out...

AI Technical Summary

Problems solved by technology

However, the mountainous glacier area covers a large amount of surface moraine, and it is difficult to distinguish the surface moraine from the surrounding features in the visible and near-infrared bands on optical remote sensing images.

At the same time, optical remote sensing images are limited by weather conditions, and it is often difficult to identify targets due to the influence of clouds and fog.

Images