Human infrared signal recognition using single PIR detector

A human recognition approach based on pyroelectric infrared (PIR) signal is proposed. The wavelet packet entropy (WPE) is employed for extracting the features of PIR signals, and then the least square support vector machine (LS-SVM) classifier is used as classifier. Relying on this approach, robust human recognition can be achieved with single PIR detector. The experiments show that the recognition rate is 91.2%, 1.3% higher than that of the traditional approach, meanwhile, the false alarm rate is reduced more than 4.0%.