Method and system for blind separation of permutation and aliasing images based on restricted Boltzmann machine

Abstract

The invention relates to a method and system for blind separation of permutation and aliasing images based on a restricted Boltzmann machine, belonging to the technical field of image processing. The present invention adopts the restricted Boltzmann machine network model to carry out fitting training on the permutation and aliasing image containing fuzzy, obtains the probability matrix by adjusting the weight, and obtains the optimal network model, and reassesses the original data set through the optimal network model. According to the feature difference between the reconstructed data set and the original data set, the fuzzy permutation aliasing area is separated. The invention uses the restricted Boltzmann machine to extract the features of the image, which can realize the automatic selection of features, greatly saves the time of feature selection, and overcomes the problems that the traditional blind separation method is not accurate and the feature domain is not easy to select. The invention can effectively separate the replacement area images for the fuzzy replacement aliasing images with different positions, sizes, numbers and noise variances of the replacement areas.

Description

technical field [0001] The invention relates to a method and system for blind separation of permutation and aliasing images based on a restricted Boltzmann machine, belonging to the technical field of image processing. Background technique [0002] Blind Source Separation (BSS), also known as Blind Signal Separation (BSS), is based only on the observed mixed output signal without requiring too much source signal and channel prior information. , the process of separating each input source signal is a research hotspot in the field of signal processing, and it is also a practical and effective signal processing method, which is widely used in image processing, data transmission, voice signal processing, mobile communication, biomedical signal processing and other fields. The blind separation of permutation-aliased images is a single-channel blind source separation problem. Different from the traditional superposition and mixed images, the replacement area and the replaced area...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method for blind separation of permutation and aliasing images based on restricted Boltzmann machines, so as to solve the problem that current blind separation algorithms cannot effectively blindly separate permutation and aliasing images of various fuzzy types; The invention also provides a blind separation system for permutation and aliasing images based on restricted Boltzmann machines

Images