A snow cover recognition method and system for synchronous satellite remote sensing sequence images

Abstract

The invention discloses a method and system for identifying snow cover of synchronous satellite remote sensing sequence images. Specifically, calculate the spectral angle and brightness difference between the multi-temporal image to be identified and the single-temporal snow cover sample, and based on the spectral angle and brightness difference, mask the multi-temporal image to be identified to generate multi-temporal highlighted features Image; calculate the dynamic time warping value, and based on the dynamic time warping value, mask the multi-temporal highlight feature image to generate a preliminary snow recognition result image; based on the snow preliminary recognition result image, use the post-processing of the classification result The method is used to obtain the snow recognition result image, which makes full use of the multi-temporal characteristics of the synchronous satellite remote sensing sequence image, and eliminates the influence of clouds without using the short-wave infrared band, and realizes the rapid and accurate recognition of snow in the synchronous satellite remote sensing image.

Description

technical field [0001] The invention relates to the field of classification and recognition of remote sensing images, in particular to a snow recognition method and system for synchronous satellite remote sensing sequence images. Background technique [0002] The spatial distribution of snow is an important input factor for hydrological and meteorological models in mountainous and seasonal snow-covered areas. The study of snow distribution and its characteristics is of great significance for atmospheric circulation analysis, climate evolution, water resource utilization and disaster analysis. [0003] Li Sanmei et al. studied the principle of snow cover identification of domestic stationary meteorological satellite FY-2C data, and proposed a snow cover identification method based on the auxiliary factor function of the threshold method, which better solved the difference in the vast area within the full disk of FY-2C. Real-time snow cover monitoring problem under underlying ...

AI Technical Summary

Problems solved by technology

However, these methods have many disadvantages: on the one hand, the preprocessing process is complicated, and different satellite data need to be geometrically corrected, atmospherically corrected, cropped, etc., and the accuracy of the geometric correction of the two data is difficult to ensure extremely high consistency

On the other hand, due to differences in spatial resolution, width, scanning area, etc., it is difficult to obtain two kinds of satellite data in the same area and at the same time, so the practicability of this method is low

Images