Pig face recognition method adopting multi-channel convolutional neural network

Abstract

The invention relates to a pig face recognition method adopting a multi-channel convolutional neural network. The pig face recognition method is technically characterized by comprising the steps of collecting a pig face data image and constructing a pig face data set for multi-channel convolutional network training; constructing a multi-channel convolutional network model: extracting eye featuresby using two shallow networks, fusing the eye features, extracting the overall image features of the pig face by using a deep network, and inputting a fusion result of the last three networks into a standard support vector machine; training a multi-channel convolutional network model and performing parameter adjustment optimization; testing the recognition performance of the network model to obtain a trained network model; and shooting and numbering the pig individuals with the face images stored in the database, and inputting shot image data into the trained network model to obtain a pig facerecognition result. According to the pig face recognition method, the image features with large individual differences are added on the basis of end-to-end, so that the robustness is high, and non-contact pig face individual recognition is achieved, and the stress response of pigs is reduced, and the automatic management level of a pig farm is improved.

Description

technical field [0001] The invention belongs to the technical field of computer vision, in particular to a pig face recognition method using a multi-channel convolutional neural network. Background technique [0002] Computer vision technology has penetrated into many fields of livestock and poultry breeding, among which the use of image processing and analysis methods to realize pig individual recognition has become a research hotspot in recent years. As an omnivorous mammal, pigs are similar to face recognition, and the facial features of different individuals are quite different. In theory, related technologies in the field of face recognition can be transferred to animal identification. Recognition has also achieved very good results. But pig face recognition has its particularity. The inbreeding characteristics of pigs lead to a high degree of similarity between pigs. At the same time, pig faces that have not been cleaned for a long time will also cover up their facial...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art, and propose a pig face recognition method using a multi-channel convolutional neural network, which solves the limitations of traditional methods and the inability of ordinary end-to-end recognition networks to strengthen learning for pig face details question

Images