Please use this identifier to cite or link to this item:
http://hdl.handle.net/10609/150314
Title: | Mejora de imágenes submarinas mediante deep learning |
Author: | Deniz Pedreira, Leroy |
Tutor: | Burguera, Antoni |
Others: | Solé-Ribalta, Albert |
Abstract: | Mobile underwater robotics has made it easier for humans to perform, and sometimes allowed humans, to carry out tasks of analysis and recognition in underwater environments. With the advanced leaps in technology in recent years, better technological resources have become accessible and a greater computing capacity has significantly improved Autonomous Underwater Vehicle (AUV), resulting in the acquisition of a vast amount of information provided by images taken by these devices. However, it is common in underwater environments that the conditions for capturing images are not optimal. As a result, there are a series of inconveniences, sometimes associated with the use of artificial light sources, ranging from excess lighting and even saturation or attenuation with distance, to loss of color depending on frequency. This work investigates techniques for automatic optimization of underwater images, including analysis, design, implementation, and evaluation. It involves the analysis of various deep learning architectures which, with the appropriate configuration and training, enable the models obtained to generate a considerable improvement in the original images. The ultimate goal is the design of a deep learning model capable of improving the quality of distorted underwater images, obtaining uniform, sharp and realistic results. |
Keywords: | image analysis underwater environments convolutional neural networks machine learning deep learning |
Document type: | info:eu-repo/semantics/masterThesis |
Issue Date: | 23-Jun-2023 |
Publication license: | http://creativecommons.org/licenses/by-nc-nd/3.0/es/ |
Appears in Collections: | Bachelor thesis, research projects, etc. |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
leroyTFM0623memoria.pdf | Memoria del TFM | 85,64 MB | Adobe PDF | View/Open |
Share:
This item is licensed under a Creative Commons License