Seybold Report ISSN: 1533-9211

Abstract

CONVOLUTIONAL NEURAL NETWORK BASED DUAL DEEP NETWORK METHOD FOR PERCEIVED DETAILS OF IMAGE AND ASGA METHOD FOR COLOR CORRECTION OF UNDERWATER IMAGE ENHANCEMENT


1M.Manju, 2Dr.S.Sukumaran, 3Dr.S.Muthumarilakshmi
1Ph.D Research Scholar, 2Associate Professor, 3Associate Professor
1,2Department of Computer Science, Erode Arts and Science College (Autonomous), Erode, Tamilnadu, India, 3S. A. Engineering College, Tamilnadu, India.
1E-Mail: manjuct21@gmail.com, 2E-Mail: prof_sukumar@yahoo.co.in, 3E-Mail: muthumarilakshmi827@gmail.com


Vol 17, No 09 ( 2022 )   |  DOI: 10.5281/zenodo.7074270   |   Licensing: CC 4.0   |   Pg no: 909-922   |   Published on: 13-09-2022



Abstract
Underwater images are causing from lightning and illumination problems. Color cast and concentrated noise is a big problem in underwater images. Traditional underwater image enhancement methods are fails to rectify these problems. For that, this work proposed Dual Deep Network method based on the Convolutional Neural Network to enhance the blurred underwater images. Reduces the color cast problems using Underwater White Balance method with color correction method of Adaptive Shades of Gray Assumption. The experimental results show that the proposed method based on the CNN has more competent and effectively enhance the underwater image than traditional State of art methods in terms of PSNR, SSIM, UCIQ and PCQI metrics.


Keywords:
Degraded Image, Transmission Map, Scene Radiance, Image Fusion, Color Correction.



Download Full Article PDF


Back to Current Issue Page