Water body extraction method based on pixel-level SAR (synthetic aperture radar) image time sequence similarity analysis

Abstract

The invention relates to a water body extraction method based on pixel-level SAR (synthetic aperture radar) image time sequence similarity analysis. The method comprises the following steps of firstly, preprocessing an SAR image data set, establishing a pixel-level SAR image time sequence through high-precision matching, and generating time sequence text data; secondly, sampling and selecting time sequences of purified water body pixels and mixed water body pixels, selecting DTW (dynamic time warping) as similarity measurement of the time sequences, and calculating a DTW value as a maximum threshold; then, calculating the DTW values of the pixel-level SAR image time sequence of all the pixels and the purified water body pixel time sequence, and segmenting an SAR image by a maximum threshold method, thus acquiring a binary image; and finally, operating the binary image so as to improve the water body identification precision by a 8-neighborhood search method. According to the method, a stable water resource distribution range can be accurately extracted, the extraction result is not influenced by hillshade, ponding in rainy season and partial vegetation, and the requirement of water body drawing can be met.

Description

technical field [0001] The invention relates to a water body extraction method of SAR images, in particular to a water body extraction method utilizing pixel-level SAR image time series similarity analysis. Background technique [0002] Water resources are an important part of natural ecological environment resources. The distribution and supply and demand of water resources have become one of the important indicators for evaluating the comprehensive competitiveness of a country or region. Remote sensing technology has been widely used in water body distribution extraction, such as MODIS, TM, SPOT and so on. In the field of microwave remote sensing, Synthetic Aperture Radar (SAR) can complete earth observation tasks without being limited by sunshine and weather conditions, and has certain penetration capabilities for certain ground objects, and has multi-polarization, multi-incident angle, multi-mode , short revisit cycle and other characteristics, it has unique advantages ...

AI Technical Summary

Problems solved by technology

[0003] However, these methods have the following disadvantages: (1) Most of these methods extract water bodies from a single image, and cannot remove the influence of rainy season water and some vegetation (such as rice); (2) Although multi-temporal water body extraction methods can be used in certain (3) It is clearly stipulated in the mapping requirements that the mapping of water bodies such as lakes and reservoirs should be based on the waterline, that is, the stable constant water level for most of the year, while the previous This method can only extract the water body distribution in the instantaneous image, but cannot extract the perennial stable water body area

Therefore, these methods cannot accurately extract the stable distribution range of water resources. At the same time, affected by mountain shadows, rainy season water and some vegetation, the extraction results cannot meet the requirements of water system mapping, and cannot be applied to the extraction and mapping of large-scale regional water body distribution. middle

Images