• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.33 no.3
• /
• pp.166-172
• /
• 1997

In order to exterminate effectively starfish, Asterias amurensis inhabited a village fishing grounds and shellfish farms on coast of Korea, Mop and sledge gear were made and sea trials for capture efficiency of starfish by each gear and towing distance were carried out by commercial dredger on the coast of Keojedo from April to May in 1995. As Starfish mop and sledge were slowly dragged over the bottom at the same time, starfish became entangled in bunches of twine and netting respectively. The gears were hauled up at intervals to remove the starfish and hand-picking operations on vessel were conducted. The results obtained are as follows : Two gears were smoothly slidden over the sea bottom and captured numerous starfish. The optimal towing distance by each gear was 300 to 500 m.The capture efficiency of starfish species by sledge was 57% compared with 43% of that by mop but mixed rate with fish or shellfish of the former was 21 times as high as that of the letter. It was concluded from sea trials that moping was effective in shellfish farms, because the mop outfit causes little damage to useful shellfishes and the mixture of starfish with fish or shellfish was low, whereas sledging can be used to clean uncultivated areas free of shellfish where starfish population is very heavy and common fishing grounds in which bottom material is rock or gravel.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.3
• /
• pp.571-580
• /
• 2022

Marine deposited waste is a major cause of problems such as a lot of damage and an increase in the estimated amount of garbage due to abandoned fishing grounds caused by ghost fishing. In this paper, we implement a real-time marine deposited waste detection artificial intelligence system to understand the actual conditions of waste fishing gear usage, distribution, loss, and recovery, and study methods for performance improvement. The system was implemented using the yolov5 model, which is an excellent performance model for real-time object detection, and the 'data screening process' and 'class segmentation' method of learning data were applied as performance improvement methods. In conclusion, the object detection results of datasets that do screen unnecessary data or do not subdivide similar items according to characteristics and uses are better than the object recognition results of unscreened datasets and datasets in which classes are subdivided.