Video Push Technology And Its Applications In Intelligent Monitoring

With the rapid development of technology and economy,video monitoring has been a common means of security surveillance in various public places.However,in most situations surveillance video can only be used in tracking after incidents because of too many surveillance video scenes to be supervised,or the carelessness of security people,or the inability of the decision makers to keep abreast of the situations on the spot.Nowadays,intelligent video surveillance systems which can promptly deal with emergencies have been urgently needed in the social public security field,and relative research and applications have been new research spot in various related fields.In this thesis,the abnormal behavior detection technology and push(PUSH)technology were applied to the intelligent video surveillance scenarios,and the intelligent video surveillance push platform was designed.The application of the abnormal behavior detection technology can help the video surveillance system avoid human negligence and misinformation and the application of the push technology changes the traditional method of getting information in common video surveillance systems.Traditional PULL technology needs active access to the videos by security people for 24 hours straight.Instead by the way of push the platform can immediately actively push the alarm notification to the intelligent terminal of the relevant persons as soon as the alarm events occur.In such platform,the extent of injury can be minimized by pushing real-timely and solving timely.The main research work in this thesis is as follows:(1)The method of crowd behavior analysis is applied to detecting abnormal behaviors.The local orderliness of the crowd motion is defined and the orderliness measurement that an individual follows the crowd around him/her to move is quantified.Meanwhile,a detection method of abnormal behavior based on the local crowd orderliness measurement has been proposed.Different from the previous methods,this thesis analyses the crowd behaviors on the macro and micro levels by employing the proposed inter-particle force model based on velocity distribution and combining with the global collectiveness descriptors.(2)Push technology of mobile terminals is analyzed and compared and the scheme that video surveillance is pushed to the mobile terminal is designed.We have made some appropriate function extensions of the applications of the existing push technology in the field of video surveillance and applied the push technology based on XMPP protocol to the security field using video surveillance.Under the scenario of monitoring,the push function is started as soon as emergencies occur and the alarm notifications and clips of the alarm video are pushed to the mobile terminals according to the association rules.As soon as a mobile terminal receives the pushed information the user will be reminded to see the alarm video by clicking the notification and then make decisions to solve it in time to avoid losses.In the meanwhile,we have devised and developed the APP on the mobile terminal which can maintain a long connection with the push module on the platform and replay the pushed video.(3)Finally,we have accomplished the detailed design of the function of the platform which can detect the abnormal behavior through the monitoring video and push the alarm video to the APP on the mobile terminals.The results of this experiment have showed that the accuracy and real features of the system good are satisfying.