Hierarchical Fish Species Detection in Real-Time Video Using YOLO
MetadataShow full item record
Original versionKalhagen, E. S. & Olsen, Ø. L. (2020) Hierarchical Fish Species Detection in Real-Time Video Using YOLO (Master's thesis). University of Agder, Grimstad
Information gathering of aquatic life is often based on time consuming methods with a foundation in video feeds. It would be beneficial to capture more information in a cost effective manner from video feeds, and video object detection has an opportunity to achieve this. Recent research has shown promising results with the use of YOLO for object detection of fish. As under-water conditions can be difficult and fish species hard to discriminate, we propose the use of a hierarchical structures in both the classification and the dataset to gain valuable information. With the use of hierarchical classification and other techniques we present YOLO Fish. YOLO Fish is a state of the art object detector on nordic fish species, with an mAP of 91.8%. For a more stable video, YOLO Fish can be used with the object tracking algorithm SORT. This results in a complete fish detector for real-time video.
Master's thesis in Information- and communication technology (IKT590)