Feeding Automation for Fish Based on Fish Movement Detection Using Computer Vision Technology

The aquaculture industry also significantly contributes to the economy by providing food protection by primary production, growing income in rural areas, and producing substantial income in export. In the aquaculture industry, though the biggest concern is fish food. Not just the significant volume...

Full description

Saved in:

Bibliographic Details
Published in:	2023 10th International Conference on ICT for Smart Society (ICISS) pp. 1 - 7
Main Authors:	Siregar, Baihaqi, Rusli, Ilham Kurnia Wahyudi, Hizriadi, Ainul, Zendrato, Niskarto
Format:	Conference Proceeding
Language:	English
Published:	IEEE 06-09-2023
Subjects:	aquaculture Automation Computer vision fish feeding Industries jetson nano Object detection Production System testing Water quality YOLOv3
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	The aquaculture industry also significantly contributes to the economy by providing food protection by primary production, growing income in rural areas, and producing substantial income in export. In the aquaculture industry, though the biggest concern is fish food. Not just the significant volume that exceeds 50 percent of the overall cost of production, but also the mechanism of feeding. In conventional aquaculture systems, fish feeding procedures still rely heavily on human labor. Negligence in the timing of fish feeding can have an impact on production performance. Efforts have been made to automate feeding such as using a scheduling system. However, in terms of performance, this method is found to be less than optimal since, under some conditions, the appetite of the fish does not agree with the fixed timetable. Automation that takes into account the feeding needs of fish is therefore needed in order to improve the effectiveness of feeding. The computer vision system is used in this research to automate the feeding of fish based on the movement of the fish towards the feeding area, which is observed using the You Only Look Once version 3 method. With this method, it is proven that the test of the fish object detection system has been implemented successfully and is effective enough to facilitate the automation system for fish feeding.
DOI:	10.1109/ICISS59129.2023.10291388