DOIONLINE 
DOIONLINE Publish In |
International Journal of Electrical, Electronics and Data Communication (IJEEDC)-IJEEDC |
 Journal Home Volume Issue |
||||||||
Issue |
Volume-8,Issue-9 ( Sep, 2020 ) | |||||||||
Paper Title |
Identification of Cough From Audio Signal using SVM Based Method | |||||||||
Author Name |
Simran H. Tamboli, Mahadev S. Patil | |||||||||
Affilition |
Student, Department of Electronics and Telecommunication, Rajarambapu Institute of Technology, Islampur, India HOD, Department of Electronics and Telecommunication, Rajarambapu Institute of Technology, Islampur, India | |||||||||
Pages |
38-42 | |||||||||
Abstract |
The proposed work presented in this paper consists of cough detection using speech processing and feature extraction scheme. Cough is the common symptom of many respiratory diseases. If cough period extends for so long time then it becomes a chronic cough and it will be so challenging task to cure. The traditional method involves self detection or doctor’s advice for detection of cough. Here the method proposed is detection of cough by audio sound. In proposed method the cough is detected by using the features of sound samples. The features detected are pitch, energy, frequency, MFCC. SVM algorithm is used for the better accuracy. The data collected is patient’s audio of cough sound and some are normal sounds. The data processed by calculating the features and then competing with the normal ones. The features are used to train the support vector machine and the performance is tested. The accuracy of detection is found better which is almost 80% and the comparative analysis against decision tree, K-nearest neighbor and discriminant analysis. Keywords - MFCC(Mel Frequency Cel Ceptrum), Feature Detection, SVM (Support Vector Machine), Audio Signal | |||||||||
| View Paper | ||||||||||
