Image de-noise process of multi-template mixed filtering

Abstract

The disclosed picture removing-noise method based on multi-template mixed filter comprises: dividing the objective picture into MXN blocks without superposition to contain every pixel in some block; defining one group of filters including the mean filter and mid-value filter, every with different removing-noise strength; defining a numerical sequence with number standing for picture information size; obtaining the noise variance of the picture, calculating the uniformity level for the picture block to thereby select one filter for finishing the removing-noise. This invention simplifies calculation, and reduces real running time.

Description

technical field [0001] The invention relates to digital image processing, in particular to an image denoising method in digital image processing. Background of the invention [0002] The existence of noise causes the image to be distorted to a certain extent. If there is too much noise, it will hinder the use of the image. At this time, the image must be denoised. The process of image denoising will inevitably cause the loss of details of the original image, so noise suppression and detail preservation are a pair of contradictions, and various denoising methods must choose a balance point between the two. Existing denoising methods have their own advantages and disadvantages, and which method to choose for denoising depends on the requirements of specific applications and the characteristics of the data used. [0003] Denoising methods can be divided into spatial domain and frequency domain according to their working domain. The existing spatial domain adaptive denoising m...

AI Technical Summary

Problems solved by technology

However, as an adaptive mean or median filter, they cannot comprehensively utilize the characteristics of the median filter and the mean filter.

Mean filtering and median filtering have their own advantages and disadvantages: mean filtering can effectively smooth the image and improve image correlation, but it is easy to blur the edges and details and cannot preserve the image structure

Median filtering can effectively remove noise and retain image details, but the image smoothing effect is not as good as mean filtering. In addition, the calculation of filtering coefficients is complicated, and the adjustment of filtering strength is limited.

In addition, most of the existing denoising methods fix the size of the filter template, and only adjust the intensity by changing the filter parameters.

In practical applications, such filtering is not only complex, but also has a limited range of filtering intensity. In some smooth areas without image details, the 3×3 template is too small and the filtering intensity is not enough, but in areas with rich details, it appears too large. , seriously impairing image detail

The frequency domain denoising method is mainly based on wavelet threshold denoising, and wavelet transform is computationally intensive and time consuming

Images