| 摘要 |
A method of automatically detect busy signal is disclosed. During the self-training process of a voice device, frequencies, thresholds of energy and on-off time parameters of the busy signal to be detected are measured. The busy signal measured is FFT-transformed, and then peak frequency of the busy signal is obtained. After several consecutive computations, parameters obtained are substituted into a busy signal detect algorithm in order to detect busy signal. This method is able to detect the busy signal of an exchange whose type is unknown, and enhances dramatically the adaptibility of a voice device to all kinds of exchanges.
|
