A NOVEL APPROACH FOR UNDERWATER IMAGE ENHANCEMENT BASED ON IMPROVED DARK CHANNEL PRIOR WITH COLOUR CORRECTION
Penulis/Author
RICARDUS ANGGI P (1); Dr.Eng. Ir. Sunu Wibirama, S.T., M.Eng., IPM. (2); Prof. Ir. Paulus Insap Santosa, M.Sc., Ph.D., IPU. (3)
Tanggal/Date
2018
Kata Kunci/Keyword
Abstrak/Abstract
The quality of underwater images is affected by characteristics of the underwater
environment, such as varying light intensity levels and varied wavelengths. Low
quality of underwater images is one of the major problems in identification of
fish species during monitoring of underwater ecosystem. Improving the quality
of underwater images is important for accurate fish identification. Some
researchers introduce various methods that address colour-correction problem for
underwater images. However, previous researches do not consider the noises
produced during the implementation of the image processing techniques. To deal
with this problem, we propose a novel method called novel contrast-adaptive
colour-correction (NCACC) to enhance the quality of underwater images that are
susceptible to bright colour distortion and various noises. The NCACC method
is a combination of an automatic level correction and a limited contrast mode of
adaptive histogram equalization method to be applied to the dark channel prior
method. As a result, we are able to improve the contrast of the images without
generating many noises. The experimental results show that the NCACC method
significantly improves the quality of the underwater images. The improvement
was assessed using the peak signal-to-noise ratio that yielded the value of 22.076
dB, which was 4.4% higher than the highest value obtained by the state-of-theart
method. We demonstrate that enhancement of underwater images is essential
to reveal the detail of underwater objects.
Rumpun Ilmu
Teknologi Informasi
Bahasa Asli/Original Language
English
Level
Internasional
Status
Dokumen Karya
No
Judul
Tipe Dokumen
Aksi
1
01_cover.pdf
[PAK] Halaman Cover
2
02_Editorial Board.pdf
[PAK] Halaman Editorial
3
03_TOC.pdf
[PAK] Daftar Isi
4
Korespondensi A Novel Approach for Und.pdf
[PAK] Bukti Korespondensi Penulis
5
05 A Novel Approach for Underwater Image Enhancement based on Improved Dark Channel Prior with Colour Correction.pdf
[PAK] Full Dokumen
6
05 Similarity A Novel Approach for Underwater Image Enhancement based on Improved Dark Channel Prior with Colour Correction.pdf