Font Size: a A A

Research And System Implementation Of Cow Face Recognition Algorithm Based On Small Samples

Posted on:2022-02-01Degree:MasterType:Thesis
Country:ChinaCandidate:S Y ChangFull Text:PDF
GTID:2493306488951199Subject:Computer technology
Abstract/Summary:PDF Full Text Request
In recent years,the intelligent breeding process of my country’s animal husbandry has been accelerating.The identification of cattle,sheep and other animals in the breeding environment has become an urgent problem to be solved.Machine learning represented by deep learning has made good achievements in the field of biometrics,and biometric technologies such as face recognition have been continuously applied to practical scenarios,providing new ideas for the identification of animals.Deep learning technology requires a large number of data samples to complete neural network model training,and the field of biometrics currently lacks large public animal data sets.In the case of a sparse sample size,the identification of animals has attracted more and more attention from researchers.This paper takes cows as a representative,conducts cow face recognition related research on cows under small sample conditions,designs a cow face recognition algorithm based on small samples and completes application examples.The main research contents are as follows:1.Make a small sample cow face data set.In this paper,in cooperation with the Sixth Dairy Farm in Pingjibao,Yinchuan City,a data acquisition system was built in the milking workshop to collect cow face video data,and related processing of the cow face video,to complete a small sample cow face data set.2.Use three methods to carry out a small sample cow face recognition experiment.This paper uses three different methods of small sample cow face recognition based on migration learning,small sample cow face recognition based on data enhancement,and small sample cow face recognition based on matching network to construct a deep network for correlation based on small sample cow face recognition Research,after experiment and comparative analysis,find a suitable solution for the development of cow face recognition system.3.Design and realize the application of a graphical interface system for cow face recognition.This paper designs a simple graphical interface system to display the entire cow identification process through a graphical interface.
Keywords/Search Tags:deep learning, small sample learning, cow face recognition
PDF Full Text Request
Related items