Underwater Image Enhancement by Independent Processing Scheme of Brightness and Chromaticity

• Published in: 2018 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) pp. 76 - 80
• Main Authors: Yoshida, Ibuki, Wong, Siaw-Lang, Paramesran, Raveendran, Taguchi, Akira
• Format: Conference Proceeding
• Language: English; Japanese
• Published: IEEE 01.11.2018
• Subjects: Asia; AWB; Brightness; chromaticity; Conferences; Degradation; DHE method; Histograms; Image color analysis; Lighting; Retinex processing; saturation histogram equalization
• ISSN: 2642-3529
• DOI: 10.1109/ISPACS.2018.8923274

Underwater images usually are blurred, low contrast and tend to be dominant by a single color component. Furthermore, when photographing, the illumination becomes uneven for the subject, thus, brightness unevenness occurs in the acquired captured images. In this paper, the underwater image is divided into brightness and chromaticity. These two components are processed independently and finally integrated to results of these two components. We apply the differential gray-levels histogram equalization (DHE) to the brightness component for the purpose of improving the contrast and sharpening. On the other hand, white balance adjustment and Retinex processing are applied to the chromaticity. White balance of color components is adjusted by the gray world (GW) method uneven optical brightness removed by Retinex. Furthermore, the chromaticity is converted to the saturation component and a colorfulness improvement processing is applied to that component. The processing result is converted to chromaticity. In the proposed method, degradation factors of underwater images are resolved and applying appropriate processing to each degradation factor. High quality improvement of the underwater image was realized.