计算机与数字工程2023，Vol.51Issue(11)：2500-2504,5.DOI:10.3969/j.issn.1672-9722.2023.11.005
基于视频实例分割的猪只监测算法
Pig Monitoring Algorithm Based on Video Instance Segmentation
摘要
Abstract
With the vigorous development of artificial intelligence,the intelligent breeding industry has become a new devel-opment prospect.Aiming at the problem of low automation and intelligence of pig farm surveillance video monitoring,a pig monitor-ing algorithm framework based on video instance segmentation is proposed,and it is named Track-YOLACT.The framework is com-posed of two parts:a segmenter and a tracker.First,segmentation is performed based on the classic single-stage instance segmenta-tion network,and then the tracker is used for tracking,so as to realize the segmentation and tracking of all pigs in the video.It pro-vides a practical framework for automated and intelligent monitoring of farms.关键词
深度学习/视频实例分割/智慧养殖业Key words
deep learning/video instance segmentation/intelligent breeding industry分类
信息技术与安全科学引用本文复制引用
张昌恒,薛辉,陈忠..基于视频实例分割的猪只监测算法[J].计算机与数字工程,2023,51(11):2500-2504,5.基金项目
民用航天十三五预先研究项目(编号:D040401-w05) （编号:D040401-w05）
国产卫星应急观测与信息支持关键技术(编号:B0302)资助. （编号:B0302）
