Design of Digital Filters for Medical Images Using Optimized Learning Based Multi⁃Level Discrete Wavelet Cascaded Convolutional Neural Network  

作　　者：Vaibhav Jain Ashutosh Datar Yogendra Kumar Jain 

机构地区：[1]Department of Electronics and Instrumentation Engineering,Rajiv Gandhi Proudyogiki Vishwavidhyalaya,Bhopal 462033,Madhya Pradesh,India [2]Department of Electronics Engineering,Samrat Ashok Technological Institute,Vidisha 464001,Madhya Pradesh,India

出　　处：《Journal of Harbin Institute of Technology(New Series)》2025年第2期55-64,共10页哈尔滨工业大学学报(英文版)

摘　　要：In digital signal processing,image enhancement or image denoising are challenging task to preserve pixel quality.There are several approaches from conventional to deep learning that are used to resolve such issues.But they still face challenges in terms of computational requirements,overfitting and generalization issues,etc.To resolve such issues,optimization algorithms provide greater control and transparency in designing digital filters for image enhancement and denoising.Therefore,this paper presented a novel denoising approach for medical applications using an Optimized Learning⁃based Multi⁃level discrete Wavelet Cascaded Convolutional Neural Network(OLMWCNN).In this approach,the optimal filter parameters are identified to preserve the image quality after denoising.The performance and efficiency of the OLMWCNN filter are evaluated,demonstrating significant progress in denoising medical images while overcoming the limitations of conventional methods.

关 键 词：digital filter image processing image enhancement OPTIMIZATION deep learning 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]