Image de-noising method based on improved two-dimensional experience modal decomposition algorithm

Abstract

The invention discloses an image de-noising method based on an improved two-dimensional experience modal decomposition algorithm. The method comprises steps of firstly, carrying out adaptive decomposition on a to-be-de-noised image through the traditional BEMD algorithm to obtain various orders of IMFs and measuring the similarity between probability density functions of the IMFs correspondingly and a probability density function of the to-be-de-noised image; then, according to a similarity measurement value, distinguishing boundary index values of noise leading modal function and a signal leading modal function; using a wavelet de-noising algorithm to carry out de-noising processing on the noise leading modal function to obtain an actual image noise; reconstructing multiple images with the same signal to noise ratio with an original image and carrying out accumulation and averaging, thereby achieving that compressing the noise in the low-order IMF; and finally, using the BEMD-DT to carry out de-nosing processing on the average image. According to the invention, by de-noising the image, de-noising effects which are better that that of wavelet de-noising and traditional BEMD de-noising method are achieved.

Description

technical field [0001] The invention relates to the field of image signal processing, in particular to an image denoising method based on an improved two-dimensional empirical mode decomposition algorithm. Background technique [0002] Image is an important visual carrier. However, in the process of image acquisition and transmission, it will be affected by external or internal factors, resulting in noise pollution. These noises reduce the quality of the image, resulting in the loss of important information in the image. Further analysis poses great difficulties. Therefore, noise reduction processing should be performed before image analysis. [0003] In recent years, some complex ultrawavelet algorithms have been applied to image denoising, such as wavelet denoising, shearlet transform denoising, curvelet transform denoising and other methods. These methods can denoise images in multiple directions. However, they need to pre-design the decomposition base. If the decomposi...

AI Technical Summary

Problems solved by technology

This solution can improve the quality of the collected pictures to a certain extent, but it increases the complexity of the hardware and costs, which is not conducive to popularization and application in practice.

Images