A new fusion feature based on convolutional neural network for pig cough recognition in field situations

作     者：Weizheng Shen Ding Tu Yanling Yin Jun Bao 

作者机构：College of Electrical and InformationNortheast Agricultural UniversityHarbinPR China College of Animal Science and TechnologyNortheast Agricultural UniversityHarbinPR China Key Laboratory of Swine Facilities EngineeringMinistry of AgricultureNortheast Agricultural UniversityHarbinPR China 

出 版 物：《Information Processing in Agriculture》 (农业信息处理（英文）)

年 卷 期：2021年第8卷第4期

页      面：573-580页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by the grant from the National Key Research and Development Program of China under Grant 2016YFD0700204-02 the Earmarked Fund for China Agriculture Research System under Grant CARS-35 the“Young Talents”Project of Northeast Agricultural University under Grant 17QC20 the University Nursing Program for Young Scholars with Creative Talents in Heilongjiang Province under Grant UNPYSCT-2020092 and UNPYSCT-2018142 the Heilongjiang Post-doctoral Subsidy Project of China under Grant LBH-Z17020 

主　　题：Pig cough recognition MFCC SVM CNN Sound classification 

摘      要：Pig cough is considered the most common clinical symptom of respiratory ***,establishing an early warning system for respiratory diseases in pigs by monitoring and identifying their cough sounds is *** this paper,we propose a new fusion feature,namely Mel-frequency cepstral coefficient-convolutional neural network(MFCC-CNN),to improve the recognition accuracy of pig *** obtained the MFCC-CNN feature by fusing multiple frames of MFCC with multiple one-layer *** used softmax and linear support vector machine(SVM)classifiers for *** tested the algorithm through field *** results reveal that the performance of classifiers using the MFCC-CNN feature was significantly better than those using the MFCC *** F1-score increased by 10.37%and 5.21%,and the cough accuracy increased by 7.21%and 3.86%for the softmax and SVM classifiers,*** also analyzed the impact of different numbers of fusion frames on the classification *** results reveal that fusing 55 and 45 adjacent frames resulted in the best performance for the softmax and SVM classifiers,*** this research,we can conclude that a system constructed by simple one-layer CNNs and SVM classifiers can demonstrate excellent performance in pig sound recognition.