Framework for Underwater Image Enhancement
| Autor: | Medha Bhat, Ujwala Patil, Zeba Patel, Uma Mudengudi, Chaitra Desai, Ramesh Ashok Tabib |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Computer science
business.industry Color balance 020206 networking & telecommunications 02 engineering and technology Sharpening Red Color Lab color space Pyramid 0202 electrical engineering electronic engineering information engineering General Earth and Planetary Sciences 020201 artificial intelligence & image processing Computer vision Artificial intelligence Pyramid (image processing) Underwater Focus (optics) business General Environmental Science |
| Zdroj: | Procedia Computer Science. 171:491-497 |
| ISSN: | 1877-0509 |
| Popis: | In this paper, we propose a framework for enhancement of underwater images. Underwater images suffer from low-contrast, blur and non-uniform illumination resulting in poor quality images. Red color in the atmospheric light is absorbed early due to its shorter wavelength, whereas the colors like blue and green penetrate deeper into water due to larger wavelength. As a result the underwater images appear bluish or greenish in color. Towards this, we propose a framework for enhancement of underwater images using color balance and Laplacian and Gaussian fusion pyramid. Here we aim to balance the color distribution of the underwater image in LAB color space, to remove the bluish-green tint caused due to atmospheric light attenuation. Focus is on sharpening the underwater image to enhance the edges distorted during the process of color balance. We emphasize on fusion of outputs obtained after color balancing and edge sharpening. We demonstrate the performance of the proposed framework using qualitative evaluation metrics and show, the results obtained through the proposed fusion framework outperforms the state of art enhancement methods. The proposed approach can also be used for enhancement of other kind of images like foggy or hazy on ground images. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|