Published Paper Details:

In the modern world underwater images have become a very important aspect of various different fields and industries. Due to this, the requirement of high-quality underwater images has become very important in recent times. These images can be used for environmental monitoring, biodiversity, documentation, ocean exploration, as well as pollution awareness. Although all of these require high-quality underwater images it has become a constant challenge in order to obtain them. This is due to the fact that underwater images tend to have various problems that they face which degrades the quality of the image itself. Some of the issues that are faced when dealing with underwater images are color distortion, low contrast and blurriness, uneven illumination, and noise from the various particulate matter that is present in the water. All these issues sort the image to make it unusable in all the fields that we previously talked about. This is why we would like to use convolutional neural networks in order to create a model that is able to enhance these images to increase the quality. In this study, we would like to address the various issues that underwater images face, and how we intend on solving them. To do this, we will have to use various data sets in order to train the model, which we will create in order for image enhancement, and then vigorously test it using various different parameters, which we will also explain in our report. Through the study, we aim to create a model, which is able to take any form of underwater image and accurately enhance it so that it can be used in the many fields that are required today., which we will also explain in our report. Through the study, we aim to create a model, which is able to take any form of underwater image and accurately enhance it so that it can be used in the many fields that require it today.