A Design of Intelligent Video Surveillance System Based on Speech Recognition Technology

SUN Da-fei,YUAN Cheng,GAO Yong
DOI: https://doi.org/10.16355/j.cnki.issn1007-9432tyut.2011.04.024
2011-01-01
Abstract:Aimed at the shortcomings of traditional video surveillance systems,a design of video monitor system was put forward,which uses speech recognition technology as an auxiliary means of monitoring.This system possessed both active early-warning and intelligent switching screen functions at the same time.Furthermore,the CMS(cepstral mean subtraction)was used to curb the impact of channel distortion on speech recognition;and the model combining and revaluation method was used to solve the training problem caused by adding updated data.The test results of experiment demonstrate that the system alarmed accurately and switched intellectually,having great values to the improvement of marketed traditional video surveillance systems.
What problem does this paper attempt to address?