Development of a Feeding Control System for Industrialized Fish Farming Based on Machine Vision
To achieve the estimation of fish population quantity and automatic feeding control in industrialized recirculating aquaculture,we focused on largemouth bass as the research subject.Based on YOLOv8,we developed a fish population estimation model and designed an automatic feeding control sys-tem integrated with feeding equipment.The system first enhanced the boundaries of the fish image using DCP operators,then employed the YOLOv8 deep learning model to recognize fish contours and classify the fish population.Combining the farmers'expertise,we set specific feed quantities for each fish population level.Finally,the system controled the feeder via PLC to dispense the predetermined amount of feed,a-chieving fully automated management of feed dispensing.Testing has confirmed that this system can accu-rately estimate the fish population quantity in real-time and automatically dispense the appropriate amount of feed through the feeding equipment.
estimation of fish quantitymachine visionYOLOv8automatic feeding