Performance comparison of different Retinex methods on enhancement of low light cave images

Authors : Bilgin Yazlık, Egemen Nazife Yazlık

Pages : 1625-1637

Doi:10.28948/ngumuh.1776861

View : 29 | Download : 54

Publication Date : 2025-10-15

Article Type : Research Paper

Abstract :Caves are spaces that open into the earth and are often either completely dark or receive very little light. Photographs taken inside caves are usually subject to extremely low light conditions and image quality is negatively affected. In this study, state-of-the-art Retinex-based image enhancement methods are used to enhance the quality of images taken in the cave. Enhanced images are evaluated both quantitatively and qualitatively. A cave image dataset consisting of 82 low-light images is used for quantitative analysis. The enhanced results from all methods are visually compared using three sample images from the dataset for qualitative analysis. Eleven different image quality assessment methods are applied for quantitative analysis. The results of the study show that Retinex-based methods are successful in the enhancement of low-light cave images. While BIMEF is recommended for applications requiring general image enhancement and structural consistency, MSRETINEX and LIME can be preferred in applications requiring high dynamic range and detail preservation. In cases where noise reduction and objective quality assessment metrics are at the forefront, SRIE provides advantages.
Keywords : Retineks, mağara, görüntü iyileştirme, performans, kıyaslama

ORIGINAL ARTICLE URL