A Hybrid Approach on Single Image Dehazing using Adaptive Gamma Correction

Putra, Oddy Virgantara and Musthafa, Aziz and Pradhana, Faisal Reza (2019) A Hybrid Approach on Single Image Dehazing using Adaptive Gamma Correction. In: The 1st International Conference On Engineering And Applied Science, 2019-08-21, Madiun.

[img] Text (Prosiding)
2019 - A Hybrid Approach on Single Image Dehazing using Adaptive Gamma Correction.pdf - Published Version

Download (565kB)
[img] Text (Reviewer)
2019 - A hybrid approach_compressed.pdf - Published Version

Download (507kB)
[img] Text (Cek Plagiasi)
A Hybrid Approach on Single Image Dehazing using Adaptive Gamma Correction.pdf - Published Version

Download (2MB)

Abstract

Since the last eruption of Mt. Kelud, a surveillance has been conducted using camera to observe the crater lake and its surroundings. The key point of this observation is based on the captured haze-free images. However, these images suffer from hazy condition because of degassing. Degassing obscures the camera vision so that the observer barely maintain to keep track of lake phenomenon. Furthermore, even many dehazing techniques have been proposed to tackle this such problems, it still leaves some other problems such as oversaturation, color distortion, and halo effects. In this paper, a hybrid method is proposed. This work incorporates two the state-of-the-art visibility restoration and contrast enhancement methods which are color attenuation prior (CAP) and adaptive gamma correction (AGC). This method is separated into three different modules. They are disparity estimation (DispE) module, transmission map enhancement (TME) module, and hazy image restoration module (ImRec). The DispE module adapts the color attenuation model for its powerful depth estimation for handling outdoor image. Subsequently, the TME module is enhanced using AGC. And the last module is calculated from the modified hazy model. The experimental results are quite impressive. It is able to maintain edge and minimize image distortion qualitatively. And the fog density measured with Fog-Aware Density Estimation (FADE) estimated at 1.448.

Item Type: Conference or Workshop Item (Paper)
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA76 Computer software
Divisions: Fakultas Sains dan Teknologi UNIDA Gontor > Teknik Informatika
Depositing User: Oddy Virgantara Putra
Date Deposited: 08 Aug 2021 15:24
Last Modified: 08 Aug 2021 15:24
URI: http://repo.unida.gontor.ac.id/id/eprint/1128

Actions (login required)

View Item View Item