Open AccessContrast Limited Adaptive Histogram Equalization for Underwater Image Matching Optimization use SURF
Author(s) -
Suharyanto Suharyanto,
Zainal A. Hasibuan,
Pulung Nurtantio Andono,
D. Pujiono,
R. I. M. Setiadi
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1803/1/012008
Subject(s) - adaptive histogram equalization , underwater , computer vision , artificial intelligence , computer science , histogram equalization , contrast (vision) , matching (statistics) , histogram , histogram matching , image quality , image histogram , turbidity , image (mathematics) , pattern recognition (psychology) , image processing , mathematics , geology , image texture , statistics , oceanography
Conditions of the underwater environment have its challenges in the underwater vision research process. Some things that make underwater imagery difficult is that light can be scattered by particles in the sea, besides that light can be absorbed by seawater, as well as the turbidity level of seawater, so special techniques are needed to get clear underwater imagery. In underwater environmental conditions, the images obtained are usually of very poor quality. Backlight and attenuation will occur this is due to water conditions, objects that dissolve easily in water, and other particulate matter so that there is the degradation of the underwater image. Because it is very important if the image is improved in quality to facilitate the process of describing objects. Image matching techniques to determine the key points of image pairs are needed in three-dimensional reconstruction research. Speeded Up Robust Features (SURF) is an image matching technique where the matching results are very dependent on the image quality. This study proposes the Contrast Limited Adaptive Histogram Equalization (CLAHE) method to increase the number of matching images with SURF. The results of the experiment showed that image matching increased by an average of 76,8 %.