Self-adaptive moving electron multiplying CCD video image denoising method

Abstract

The invention discloses a self-adaptive moving electron multiplying CCD video image denoising method. According to the method, bidirectional registering is performed on the reference frame and the current frame of a low-light video sequence by adopting a determinant projection mode, the sum of deviation squares of projection of two sequentially continuous frames of images is calculated and the obtained function of the sum of deviation squares is a function related to displacement, and then the maximum displacement of the two sequentially continuous frames of images is found by traversal of the matrix of the sum of deviation squares; and the overlapped area of the two frames is obtained by the calculated displacement coordinates of the two sequentially continuous frames of images, image information of the overlapped area is not changed after the overlapped area is obtained and noise randomly changes and meets a Gaussian noise model, the displacement direction of the images is judged, bidirectional registering is performed on the images, and then weight filtering is performed to remove noise. Noise of electron multiplying CCD moving videos can be removed, and the image correction effect is great so that a ghost problem caused by the conventional moving video denoising method can be avoided.

Description

technical field [0001] The invention belongs to the field of denoising of electronic multiplication CCD system imaging, in particular to a method for self-adaptive denoising of moving electronic multiplication CCD video images. Background technique [0002] With the development of science and technology, the world has entered the era of photons. With the rapid development of information technology, the means for people to obtain information are expanding to different bands and wider fields. Image information, as the most direct and effective information, has been widely valued by the world. How to obtain clearer and more accurate image information has become an important issue for various countries. Low-light imaging technology conforms to this development trend and has become one of the emerging high-tech technologies for both military and civilian purposes that developed countries in the world are vigorously developing. [0003] Low-light imaging extends the visual rang...

AI Technical Summary

Problems solved by technology

Image processing is a fast, effective, and low-cost method to remove image noise. At present, traditional image denoising algorithms can only achieve certain effects on still image denoising (Liu Han, Liang Lili, Huang Lingshuai, "Blocking Singular Value Decomposed two-level image denoising algorithm" "Acta Automatica Sinica", Vol. 42, No. 2, February 2015), but it cannot be applied to shooting low-light moving image videos, and the popular wavelet transform method is computationally complex. can't be applied in practice

Images