Background
Fisheries is one of the resources which Oman is wealth of and considered as one of the important aspects of Omani economic diversification. Natural coral reefs are important to sustain the fisheries ecosystem in Oman. However, it is currently suffering from predation by the crown of thorns starfish (an animal that feed on coral), storm damage, diseases, fluctuating temperature and coral bleaching. Therefore, artificial reefs have been used in the rehabilitation of the damaged natural reefs. Thus, different projects have been done in different regions and millions Omani Rials are spent. The artificial reef sites need regular monitoring which is currently done manually by sending divers to the coral sites then observe and record the changes. Further, this method is affected by lack of trained and qualified divers and the depth they can reach. The cost of diving equipment is also high, and divers are often at risk. Because of the importance of the artificial coral reefs, there is a need to facilitate and automate the monitoring process in terms of efficiency and cost.
This project aims to develop an automated process for artificial coral reef monitoring. Underwater drones will be used to acquire frames for the site, then a software that analyze the recorded video will provide the fish species which are attracted to the site and the quantity of each kind. The proposed system will be able to monitor the artificial coral reefs sites by observing the biological changes in the site using low cost, short time, and efficient process.
Data Collection and Pre-processing
Training DL model
Testing