基于FaceNet的犬只鼻纹识别  

Canine nose print recognition based on FaceNet

作　　者：李梦晗 张晨蕾 田存伟[1] Li Menghan;Zhang Chenlei;Tian Cunwei(School of Physical Science and Information Engineering,Liaocheng University,Liaocheng 252000,China)

机构地区：[1]聊城大学物理科学与信息工程学院,聊城252000

出　　处：《现代计算机》2024年第4期9-15,28,共8页Modern Computer

摘　　要：随着养宠家庭的增加,犬只身份识别成为一个重要问题,犬鼻纹独一无二,可以用来准确地辨认犬只身份。为了实现犬鼻纹识别,设计了一种基于FaceNet网络的犬只鼻纹识别方法。FaceNet的Backbone采用ResNet18,损失函数采用Tripletloss。并通过SRGAN网络对犬鼻纹图像预处理,放大图像分辨率,使鼻纹图像更加清晰。基于FaceNet网络的犬只鼻纹识别方法为犬只身份识别和鉴定提供了一种有效可行的解决方案。With the increasing number of pet⁃owning households,canine identity recognition has become an important issue.Canine nose prints are unique,and can be used to accurately identify canine identities.In order to achieve canine nose print recog⁃nition,designed a method based on the FaceNet network.The Backbone of FaceNet utilizes ResNet18,and the loss function em⁃ploys Triplet loss.Additionally,SRGAN network preprocessing is applied to canine nose print images to enhance image resolution,making the nose print images clearer.Based on the FaceNet network offers an effective and feasible solution for canine identity rec⁃ognition and authentication.

关 键 词：FaceNet 犬鼻纹识别 ResNet18 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S829.2[自动化与计算机技术—计算机科学与技术]