Multispectral remote sensing image change detection method and system based on deep learning

Abstract

The invention discloses a multispectral remote sensing image change detection method and system based on deep learning, and belongs to the technical field of image processing. The method comprises thesteps: calculating the change vector amplitude of a remote sensing image; obtaining a pseudo training sample set according to a change vector amplitude EM algorithm, wherein the pseudo training sample set comprises a mark sample set (including a change type sample set and a non-change type sample set) and a non-mark sample set; constructing two networks, namely a student network and a teacher network, constructing a cross entropy loss function for the mark sample set, and constructing a mean square error function for the non-mark sample set; optimizing the student network by adopting a stochastic gradient descent optimization algorithm, and updating weight parameters of the teacher network in each training round; and obtaining a corresponding final change detection result according to thefinal teacher network. Besides, the non-mark sample set is added in network training to participate in training so that the change detection result is enabled to be more reliable and more robust.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a deep learning-based multispectral remote sensing image change detection method and system. Background technique [0002] Change detection of remote sensing images is to quantitatively analyze and determine the characteristics and process of surface changes from remote sensing data of different periods. Scholars from various countries have proposed many effective detection algorithms from different angles and applied research. Generally speaking, according to whether training samples are needed in the detection process, change detection can be divided into three categories: unsupervised change detection algorithms, semi-supervised change detection algorithms Detection algorithms and supervised change detection algorithms. Because unsupervised change detection algorithms do not require training samples, and the modeling process does not require prior knowled...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that the training of the network is prone to mode collapse

Images