Abstract
The cry sound produced by an infant is the result of his or her physical and psychological condition and/or internal or external stimuli. It has been shown that different cry origins, such as pain, hunger, insult and fear, exhibit different cry patterns. Different type of cry will produced different sound such as pitch and length of silent voiced. There are many methods to use for extract the information where we can use Mel-frequency cepstral coefficients (MFCC) and linear prediction cepstral coefficients (LPCC). For this thesis, the LPCC was chosen because over more intuitive frequency domain representations were basically two: applicability and ease of use. The main objective is to build up a characteristic that can be use for recognition phase for automatic infant cry recognition system. There nine sample where three type of infant in three different conditions was use. Each sample was test for pitch with three method and voiced/unvoiced detection. The result from this four method shown different sample will produce different characteristic. Most of method use amplitude and frequency to differentiate the feature.
Metadata
Item Type: | Thesis (Degree) |
---|---|
Creators: | Creators Email / ID Num. Abdullah, Muhammad Hafiz UNSPECIFIED |
Contributors: | Contribution Name Email / ID Num. Thesis advisor Yasmin, Farah UNSPECIFIED |
Subjects: | T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Electronics > Applications of electronics T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Electronics > Pattern recognition systems |
Divisions: | Universiti Teknologi MARA, Shah Alam > Faculty of Electrical Engineering |
Programme: | Bachelor in Electrical Engineering (Hons.) |
Keywords: | Role, individual, assessment |
Date: | 2009 |
URI: | https://ir.uitm.edu.my/id/eprint/67692 |
Download
67692.PDF
Download (232kB)