Automated underwater image analysis
Limited data availability in deep sea habitats due to acquisition struggles is a historic issue, driven primarily by the vastness and isolation of this environment. This lack of data significantly impedes ecological research. Increasing use of autonomous equipment such as Autonomous Underwater Vehicles (AUVs) is helping to resolve this issue, but in the case of image data, labour-intensive analysis creates a severe problem. The development of artificial intelligence to conduct such tasks, is therefore necessary to optimize this process and cope with accumulating data. In collaboration with Gardline and Heriott-Watt University, my PhD will aim to investigate automating the analysis of deep-sea benthic images using computer vision techniques.