Research On The Extration Algorithm Of Physiological Parameters In Polygraph

Lie-detection is used at assisting investigation and case solving and becomes an important means to cracking down on crimes and crime prevention. Especially with the modern polygraph, the measurement of the mental physiological parameter is becoming accurate.The psychological pressure arising from lying usually results in rapid heartbeat, blood pressure rising, palm sweating, body temperature rising, muscle fluttering, respiratory capacity or respiratory rate being abnormal and so on.Based on the principles above, in this paper, the changes of skin resistance, blood pressure, blood volume,heart beat at various psychological pressure are studied at first. Then the physiological parameter index of polygraph is provided which relies on them. The practical application has proved its accuracy.Alternatively, human's muscle fluttering at nervous psychology can be transmitted through voice and results in the alteration of speech signal. After a lot of experiments and testing research, the frequency band related with lie-detection is found. To decrease the impact of noise in speech signal and increase the test accuracy, this paper puts forward an improved wavelet threshold function, and combines it with voiced-unvoiced distinction so as to improve the method of noise reduction with wavelet threshold. The simulation experiment shows that this method can extract the original speech signal more effectively. Lastly, in this paper, the process of the extraction algorithm of voice physiological parameter index is put forward and it is revealed by the simulation experiments that it can be used as powerful auxiliary parameter to the polygraph.