Towards automatic hole detection of a net for fish farms by means of robotic intelligence
In the last decades fish farms became one of the most important sources of seafood. This industry is facing complex and costly problems like net holes, especially due to unexpected situations, such as depredators and storm effects. This is a complex problem because fishes can escape from the fish fa...
| Autores: | , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2023 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/391368 |
| Acceso en línea: | https://hdl.handle.net/2117/391368 |
| Access Level: | acceso abierto |
| Palabra clave: | Robot vision Underwater robots Computer vision Underwater manipulation Aquaculture Visió artificial (Robòtica) Àrees temàtiques de la UPC::Informàtica::Robòtica |
| Sumario: | In the last decades fish farms became one of the most important sources of seafood. This industry is facing complex and costly problems like net holes, especially due to unexpected situations, such as depredators and storm effects. This is a complex problem because fishes can escape from the fish farms containers or a depredator can enter in the container. To solve this problem divers are needed, but this solution is difficult and sometimes can be dangerous for the diver. The main objective of this work is to present the current state of a system where an underwater robot can detect holes in the net of a fish farm. Once the robot detects the hole it will proceed to manipulate it. This task is bordered using convolutional neural networks and the BlueROV2 platform with the Newton Gripper from BlueRobotics, which will be upgraded in a second stage to perform preliminary net repairs. This work contributes in the area of aquaculture, computer vision, underwater inspection and manipulation. |
|---|