摘要
动物身份识别是进行动物管理、动物保护与生态学研究的基础。随着技术的进步,动物身份识别的方法也不断丰富,而基于声纹技术的动物身份识别在识别距离、识别准确率、系统运行成本及环境适应性等方面拥有明显的优势,是动物身份识别领域的研究热点,其研究内容主要集中在动物物种识别、个体识别和声音状态识别等方面。目前,在实际应用中还存在复杂场景中识别精度和稳定性不够、数据采集难度大、缺乏标准数据集等问题,需要探讨相应的解决方法。
Animal identity recognition is the basis of animal management,animal protection and ecology research.With the progress of technology,the animal identity recognition methods also enrich,and based on voice print technology,the animal identity recognition in recognition distance,recognition accuracy,system operation cost and environmental adaptability has obvious advantages,is a research hotspot in the field of animal identity recognition,its research content mainly focuses on animal species identification,individual recognition and sound state recognition,etc.At present,there are still some problems in practical application,such as insufficient identification accuracy and stability,difficulty in data acquisition,and lack of standard data sets in complex scenarios,so corresponding solutions need to be discussed.
作者
南兆营
喻洪江
孔庆波
NAN Zhao-ying;YU Hong-jiang;KONG Qing-bo(Criminal Investigation Police University of China,Shenyang,Liaoning,110034,China;Chongqing Police College,Chongqing,401331,China)
出处
《动物医学进展》
北大核心
2025年第7期115-119,共5页
Progress In Veterinary Medicine
基金
重庆市教委科学技术研究重大项目(KJZD-M202001701)。
关键词
声纹技术
动物身份识别
深度学习
Voice print technology
Animal identity recognition
Deep learning
作者简介
南兆营(1980-),男,山东梁山人,副教授,硕士,主要从事声像资料检验的教学、科研工作。;通讯作者:孔庆波。
