Underwater image enhancement

• Main Author: Kua, Bella Jing Xuan
• Other Authors: Chau Lap Pui
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/70879
• Institution: Nanyang Technological University

The ocean is very vast and it makes up seventy percent of the Earth’s surface yet, less than five percent of the ocean have been discovered and explored. In this technology inclined generation, inventions and creations are created to enable humans to take images and footage underwater. The purpose of capturing and recording of underwater images are endless, ranging from research and study purposes to leisure uses or even tracking an animal species. Hence, the quality of the underwater images captured is very essential. Due to different light conditions in the air and in the waters, the quality of underwater images is flawed. It suffers from low contrast, lack of visibility, color distortion etc, Moreover, due to the existence of dust-like particles in the water, underwater images also suffer from the backscatter effects in various degrees. Hence, this project aims to design and implement an image processing method to improve contrast and increase the visibility of an image to improve the quality of underwater image. There are many image enhancement and restoration methods to enhance underwater images. Since there are plentiful image prior models based on haze removal, this can be applied to remove haze-like effects evident on underwater images. Moreover, due to color distortion and low illumination observed in enhanced underwater images, color-balancing algorithm would be implemented as well to enhance the overall quality of underwater images. In this project, the characteristics of underwater images would be carefully analyzed and the proposed method which includes both dark channel prior (DCP) and color balancing algorithm is recommended to enhance the overall clarity of underwater image. The works of this project are attainable and coded in the MatLAB software programme.