Method and device for unsupervised change detection based on full polarization time-series SAR images

Abstract

The non-supervised change detection device based on full-polarization time-series SAR images of the present invention includes: an image preprocessing unit, which registers and filters a pair of full-polarization time-series SAR images of different time phases in the same geographical range; preliminary difference image The generation unit performs hypothesis testing on the corresponding positions in the pair of full polarization time-series SAR images to generate a preliminary difference image; the optimal threshold selection unit uses the KI algorithm to select the optimal threshold for the generated preliminary difference image and a detection result generating unit that uses the selected optimal threshold to perform binary segmentation on the preliminary difference image, assigning a value indicating no change to pixels smaller than the threshold, and assigning a value indicating change to pixels greater than the threshold value, and finally generate the detection result. According to the present invention, full-polarization information of SAR images can be fully utilized to realize automatic and high-precision extraction of change information from SAR images of different time phases.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, and relates to a method for non-supervised change detection of time-series images, in particular to using the hypothesis test of the Wishart distribution conforming to full polarization information and the KI minimum classification error method to detect time-series synthetic aperture radar (Synthetic Aperture Radar) A method and system for unsupervised change detection on Aperture Radar (SAR) images. Background technique [0002] Synthetic Aperture Radar (SAR) technology is an important means of earth observation. Compared with optical remote sensing technology, Synthetic Aperture Radar (SAR) system has the advantage of all-weather and all-weather acquisition of remote sensing data. . At the same time, SAR is also very sensitive to the geometric and physical characteristics of the detection target, and is a good source of change detection information. Therefore, it is wi...

AI Technical Summary

Problems solved by technology

However, existing change detection algorithms have problems such as insufficient utilization of polarization information and the need for manual intervention.

[0004] Moreover, in a complex geographical environment, the backscattering characteristics of ground objects are easily affected by many factors: such as the biology of ground objects, physical parameters, dielectric constant, environmental factors (such as weather conditions), radar parameters (radar wave Incidence angle, wavelength, polarization mode), etc., so even for unchanged ground objects, there are different degrees of differences in the gray value and statistical characteristics of SAR images in different phases. Complexity and Difficulty of Change Detection Techniques

Images