The research of target auto-sensing system based on energy difference analysis

Abstract

At the environmental awareness in the level of the Internet of Things, the target sensing system gradually becomes an important technology. Therefore, some aspects are becoming the key to improve the quality of applications of the Internet of Things, such as the pretreatment, feature extraction, classifier and hardware realization and so on. In this study, on the basis of analysis of the specific noise source generation mechanism, different testing algorithms are used to carry out the acoustic signal pretreatment, feature extraction and classifier design, then testing algorithms used in each process are integrated to gain the best result. Ultimately, the specific target automatic detection system has been achieved, including adaptive filtering algorithms, wavelet energy feature, and similarity coefficient clustering algorithm.