Open AccessAn efficient residual learning deep convolutional neural network for de-noising medical images
Author(s) -
Heren G. Chellam
Publication year - 2022
Publication title -
international journal of health sciences (ijhs) (en línea)
Language(s) - English
Resource type - Journals
ISSN - 2550-696X
DOI - 10.53730/ijhs.v6ns3.6073
Subject(s) - convolutional neural network , artificial intelligence , residual , computer science , noise reduction , pattern recognition (psychology) , deep learning , convolution (computer science) , noise (video) , peak signal to noise ratio , artificial neural network , mean squared error , image (mathematics) , algorithm , mathematics , statistics
Image denoising is a pre-processing technique that is done in every image processing applications. It plays a significant role in the performance of any methods. The objective of this paper is to remove Gaussian noises at different noise levels in medical images. This paper proposed an efficient Deep Convolution Neural Network model for denoising medical images to remove Gaussian noise using Residual Learning. Convolutional Neural Networks are a class of deep neural networks that can be trained on large databases and have excellent performance on image denoising. Residual learning and batch normalisation are various techniques used to enhance the training process and denoising performance. The proposed RL-DCNN method is tested with 20 layers and evaluated using the performance metrics Peak Signal to Noise Ratio, Mean Square Error and Structural Similarity. It is compared with Denoising Convolutional Neural Network and Shrinkage Convolutional Neural Network models and proved to be better than the other methods.