Please use this identifier to cite or link to this item:
http://nopr.niscpr.res.in/handle/123456789/63131
metadata.dc.identifier.doi: | https://doi.org/10.56042/jsir.v82i12.5132 |
Title: | DCNN-HBA: Honey Badger Optimization and Deep Convolutional Neural Network Based a Novel Hybrid Model for Producing Quality Image |
Authors: | Niu, Sihan Singh, Vineeta Kumar, Alok Verma, Deepak Kumar Kumar, Sunil Kaushik, Vandana Dixit Chen, Zhiliang Joshi, Kapil |
Keywords: | DCNN;Median filter;Medical imaging;Multi-focus image;Noise removal |
Issue Date: | Dec-2023 |
Publisher: | NIScPR-CSIR, India |
Abstract: | The processing of images is a major task in several domains like medical treatment, military, and surveillance. However, the major reasons, like environmental criteria and technical issues made the imperative information tainted. The blurriness represents degradations induced on the image that affected image contrast. There exist several techniques based on image enhancement to improve image quality, but most of these techniques are complex to examine and impose image degradation. An optimized deep technique is devised for producing quality pictures in which the input image is gathered from the database. The pre-processing is done utilizing the median filter to discard the artefacts as well as the noise accumulated in the images. The image enhancement is done with a Deep Convolutional Neural network (DCNN) and the weight update in DCNN is carried out with the Honey Badger Optimization Algorithm (HBA). Thus, the DCNN-HBA helps to enhance the quality of the image without any kind of degradation, like blurriness. The DCNN-HBA technique provides better results with the highest mutual information (MI), highest universal quality index (UQI), maximum UQI, and enhanced efficacy of image enhancement. The highest structural similarity index measurement (SSIM) is the maximum SSIM. |
Page(s): | 1304-1315 |
ISSN: | 0022-4456 (Print); 0975-1084 (Online) |
Appears in Collections: | JSIR Vol.82(12) [December 2023] |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
JSIR (82)12 1304-1315.pdf | 9.07 MB | Adobe PDF | View/Open |
Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.