基于计算机视觉的鱼类智能投喂方法研究进展

打开文本图片集

中图分类号：S126 文献标识码：A 文章编号：2095-5553（2025）07-0198-08

Abstract：Feedingisoneofthemain tasks inaquaculture，andhowtoreducethecostoffeedingisthekeypointof maximizing theprofitsofaquaculture.Usingcomputer vision technologytomonitor the feeding behaviorof fishand quantifythe intensityoffish feding desirecan realize automatic feeding on demand，reduce feed waste，and improve feed utilization.This paperreviews theresearch progressoffour kindsoffish intellgentfeeding methods based oncomputer vision，such as food detection，optical flow method，texture and deep learning，and analyzes theadvantages and disadvantagesof each method.Bait detectionmethodissimpleandeasytoimplement，butitisdificulttoidentify theresidualbaitacurately.Opticalflowmethodcancapturefishmovementinformationefectively，butitis easilyaffected by environment and light. Texture and otherfeature methodsuse more types offeatures and more efective information， buttheyare not suitable for high-densityculture.Deep learning method has high recognitionaccuracy，strong robustness， largecalculation amount，and high requirements for equipment computing power.Basedon this，three research directions of large-scale data set，efficient lightweight depth model and “Internet of Things +，， intelligent feeding are proposed. It provides reference for further improving the maturity and practicability of intelligent feding method.

Keywords：fish；computer vision；feeding behavior；feed detection；optical flow；deep learning；intelligent feeding

0 引言

我国是世界第一水产养殖大国，水产养殖产量连续多年排名世界首位。（剩余19587字）