利用机器学习开展畜禽品种鉴定的研究进展

打开文本图片集

中图分类号：S813.1 文献标志码：A 文章编号：0366-6964（2026）02-0582-13

Research Progress in Livestock and Poultry Breed Identification Using Machine Learning

QIU Danni1，LI Yipeng1， ZHAO Shanghui'，WANG Ruiqi¹，LIU Xiaohong²，QIU Xiaotian 3* YUAN Xiaolongl*

（1.National Key Laboratory of Swine and Poultry Breeding，National Engineering Research Center for Breeding Swine Industry， Guangdong Provincial Key Laboratory of Agricultural Animal Genomics and Molecular Breeding，College of Animal Science， South China Agricultural University，Guangzhou 510642， China; 2. School ofLife Sciences，Sun Yat-sen University，Guangzhou 51O275，China; 3.National Animal Husbandry Seroice，BeijinglOol25，China）

Abstract：Breed identification is fundamental to livestock conservation and plays a vital role in genetic resource management， breed purity assessment， breeding strategy implementation， product traceability，supply chain optimization，and ensuring food and biosafety. This review summarizes both traditional and modern methods for breed identification of common livestock and poultry（pig， catle，sheep，dog，chicken），and systematically reviews recent advances in machine learning-based breed identification research.Traditional methods，such as morphological assessment and physicochemical indicator analysis in animals，are increasingly inadequate for large-scale breed identification due to their limited accuracy，operational complexity，and time-consuming nature. With advancements in intelligent animal husbandry and high-throughput omics technologies，machine learning-based approaches have gained wide application. This paper reviews major achievements in bred identification using machine learning，analyzes the strengths and limitations of various algorithms，and highlights challenges in practical use. It offers insights for future development and supports molecular breeding and precision livestock management.

Keywords： machine learning； breed identification; livestock and poultry； SNP； image recognition;omics；intelligent animal husbandry

∗ Corresponding authors： YUAN Xiaolong，E-mail： yxl@scau. edu. cn；QIU Xiaotian， E-mail：qxt-nahs@agri. gov. cn

我国地域辽阔，疆域横跨多个气候带，地理气候的多样性使中国成为畜禽遗传资源最丰富的国家之一，中国畜禽遗传资源约占世界总量的六分之_[1]。（剩余22735字）