SMART FISHING SYSTEM FOR CATCHING SELECTIVITY AND OPTIMIZATION OF COASTAL FISHERMEN'S CATCH

Abstract

This invention concerns an intelligent system for fishing selectivity and catch optimization of coastal fishermen designed to improve fishing efficiency in an environmentally friendly manner. This system consists of an environmentally friendly fish trap equipped with an underwater camera, aquatic environmental sensors, an underwater lighting module, an edge computing-based microcomputer unit, a communication module, and a power management module. Fish images and environmental data are automatically processed using the EcoTrapNet-DNN algorithm, which is a multi-task deep neural network based on underwater acquisition parameters configured to perform image quality validation, image quality improvement, fish detection and segmentation, size estimation based on geometry calibration, classification of fish types and sizes, and determination of catch selectivity scores.Based on the analysis, the system provides recommendations for fishing selectivity or, in certain embodiments, controls the selective mechanism in the fish trap so that target fish are retained, while non-target or substandard fish are released. This system can reduce bycatch, improve the quality and value of the catch, and support sustainable fishing practices for small- to medium-scale coastal fishers.

Images