A feature extraction method for abnormal sounds in public places

Abstract

The invention relates to a method for extracting characteristics of abnormal noise in public places and belongs to the technical field of audio signal processing. The method comprises utilizing T distribution based random noise to replace gaussian white noise in the original ensemble local mean decomposition model to deduce a TD-ELMD (Based on ensemble local mean decomposition) model; performing decomposition on abnormal noise signals in the public places through the TD-ELMD model to obtain a series of PFi (Product Functions) from low frequency to high frequency; utilizing the ratio of the energy of the orders of the PFi and the total energy of the original signal as a characteristic vector of the abnormal noise in the public places; inputting the characteristic vector into an SVM (Support Vector Machine) classifier to be classified and identified. Compared with the currently common extraction method, the characteristics of the abnormal noise signals in the public places can be reflected by the TD-ELMD model well, the characteristic description capability is strong, and the identification on the abnormal noise in the public places is good in robustness.

Description

technical field [0001] The invention belongs to the technical field of audio signal feature extraction and pattern recognition. Background technique [0002] Public places refer to various places where the public engages in social life, such as squares, stations, schools, etc. All kinds of abnormal events and crimes in public places are closely related to people's life, property and safety. At present, the security monitoring of public places is mainly based on the video monitoring platform, and most of them do not have the audio monitoring function. Abnormal events in most public places are accompanied by abnormal sounds such as gunshots, explosions, broken glass, screams, etc. The production of unusual sounds indicates that anomalous events have occurred or are about to occur. Making up for the lack of video surveillance platform through audio surveillance technology has become the development direction of security surveillance in public places. Due to the lack of rele...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to propose a feature extraction method suitable for abnormal sounds in public places, aiming to solve the key theoretical and technical problems in the feature extraction of abnormal sounds in public places, that is, the first is the problem of modeling background noise in public places; Feature Extraction of Abnormal Sounds in Places

To solve the core problem of feature extraction of abnormal sounds in public places, it is first necessary to obtain the distribution model of background noise in public places

Images