Modified Multi-Wavelet Noise Filtering Algorithm for Mammographic Image Denoising Using Recurrent Neural Network

Abstract

The digital mammographic images are affected by several types of noises which require filters to denoise the noise level This will help the medical practitioner to enhance the image quality of the mammograms and helps them in giving accurate diagnosis There are so many works on image denoising technique but there are not much which gives emphasis on the mammographic images In application point of view medical images are classified as Multispectral Image used for satellite surveillance RGB standard colour scheme Image or other digital versions of the film image i e in our case its mammographic image For every image type it requires different approach for denoising because in each type of image it contains different factors in it In denoising the mammographic image the filtering technique that is to be applied depend on its noises at each resolution level of the microns to make the micro-classification of the cancerous tissues to that of the bright water dense patches caused by the calcium salts in the mammary glands Thus any single algorithm cannot provide similar performance range for different types of noise because not every method is effective for the scenario of mammographic image denoising In the given study we have shown a method for the mammographic image denoising which is having higher accuracy and the performance range is suited for denoising applications raphic image denoising