Improved techniques to remove color cast and contrast enhancement for underwater images / Wong Siaw Lang
Light scattering and absorption of light in water cause underwater images to be poorly contrasted, hazy and dominated by a single color cast. The Gray World (GW) method has been widely reported to enhance images but not in underwater images. This is because GW method introduces red artifacts in unde...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2019
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/11117/1/Wong_Siaw_Lang.pdf http://studentsrepo.um.edu.my/11117/2/Wong_Siaw_Lang.pdf http://studentsrepo.um.edu.my/11117/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Light scattering and absorption of light in water cause underwater images to be poorly contrasted, hazy and dominated by a single color cast. The Gray World (GW) method has been widely reported to enhance images but not in underwater images. This is because GW method introduces red artifacts in underwater images. In this study, an Adaptive Gray World (AGW) method is introduced where apart from computing the global mean, the local mean of each channel of an image is taken into consideration and both are weighted before combining them. However, the AGW method can only remove the color cast and reduces the red artifacts, but does not enhance the contrast of the image. Thus, two enhancement methods, Differential Gray-Levels Histogram Equalization (DHE) and Differential Gray-Levels Histogram Equalization for Color Images (DHECI), are used to improve the contrast of the underwater images. In this thesis, two methods are proposed. First, a novel parallel structure consisting of AGW and DHE method is proposed to deal with the chromaticity and intensity components separately. The outputs of both chromaticity components of AGW and intensity components of DHE are fused to form the enhanced image. Widely-used database underwater images are used to validate the effectiveness of the proposed method. The results of the proposed parallel structure method are compared with the state-of-the-art methods using qualitative and quantitative measures. The quantitative measures include Entropy, Patch-based Contrast Quality Index (PCQI), Underwater Color Image Quality Evaluation (UCIQE) and Underwater Image Quality Measure (UIQM). The proposed parallel structure method gave better visual quality and in most cases produced better quantitative scores when compared to the state-of-the-art methods. Next, the second proposed method uses a sequential structure that consists of AGW and DHECI. In this sequential structure, before the color cast is removed using AGW method, the red and blue channels of the underwater image are first compensated. The output image of the AGW is then transformed from Red Green Blue (RGB) to Hue Saturation Intensity (HSI) color space to facilitate the color image enhancement to take place. DHECI method is adopted in this sequential structure as it can improve the contrast and colorfulness of the underwater image. This proposed sequential structure method is also validated with the same dataset of underwater images and compared with the same state-of-the-art methods. The aforementioned quantitative assessments used in assessing the proposed parallel structure are also used in this sequential structure. The results showed that the proposed sequential method performed better when compared to the state-of-the-art methods.
|
|---|