Research And Implementation Of Security Systems Based On Motion Detection

In recent years, with the enhancement of safety awareness,traditional video surveillance systemhas been unable to meet the people’s actively request of defense againstpotential danger. With the rise ofintelligent hardware, more intelligence is the inevitable trend of development.This thesisstudies and designsa security system based on motion detection. The purpose is to make the system be able to take preventivemeasures.This thesis references to a large number of related literature,and then designs the system byanalysis the actual situation. We focus on two core technologies: video de-noising and movingdetection.First, several of the existing algorithms are analyzed and compared,then an adaptive bilateralfiltering algorithm based on the characteristics of the monitoring system are proposed.The method canachieve better de-noising effect without using multi-frame image.In the other hand,we proposed a newdetection method which combined Gaussian mixture model and average background model.In softwaredesign, the design and implementation of the functional modules required for the system are carried out.The system used in the development environment and tools used in simulation and hardware componentsare introduced.Finally, the simulation process is completed and tested. Experimental results show that thismonitoring system is able to meet the actual needs of the video surveillance, is reliable and easy to operate,meanwhile it has a good economic benefits.