ACADSTAFF UGM
| Judul/Title | Multiresolution Modified Grey Level Difference for Respiratory Sound Classification |
| Penulis/Author | Prof. Dr. Ir. Risanuri Hidayat, M.Sc., IPM. (1); Prof. Ir. Hanung Adi Nugroho, S.T., M.Eng., Ph.D., IPM., SMIEEE. (2) |
| Tanggal/Date | 2016 |
| Kata Kunci/Keyword | |
| Abstrak/Abstract | Changes in the 1D signal amplitude is comparable to the texture of the image. In this study, we modified texture analysis techniques and used it for lung sounds feature extraction. Grey level difference (GLD) is applied to the lung sound subbands result of discrete wavelet transform (DWT) and wavelet packet decomposition (WPD). Lung sounds are considered as the image with 1xN dimension and scaled to 0-255 value as the 8bit image. The lung sound GLD’s features are classified using a multilayer perceptron neural network (MLP-NN) with 3-fold cross validation. The experiment results showed that the highest accuracy up to 100% with the lowest accuracy of 90% for the five classes of the lung sound data. MLP combination of N-15-5 and Db8 provide the best results for DWT or WPD. |
| Rumpun Ilmu | Teknik Elektro |
| Level | Internasional |
| Status |
| No | Judul | Tipe Dokumen | Aksi |
|---|