A Design & Research Of Video Surveillance System For Mobile Robot

In the last decade,with the developing of information technology,intelligent robots have been applied to human production on a large scale.The mobile robot with autonomous navigation and positioning functions can be deployed in various places.At present,the research of surveillance video acquisition system based on mobile robot has become a hot topic in academia and industry.This paper mainly solves the engineering problem of applying video surveillance technology to the actual mobile robot working environment.First of all,this paper builds a mobile robot surveillance video acquisition platform.Combined with the existing robot chassis technology,this paper designed the corresponding hardware control unit and the upper layer equipment management unit,to achieve the robot equipment data interaction and device controlling.Then,this paper studies the video acquisition and coding method.In order to ensure the low power consumption,the platform hardware acceleration ability is used to encode video data.At the same time,the software scheme supports various coding methods,making it possible for future developing.The module runs on the general hardware equipment and software platform,which has good equipment replacement ability and software transplantation potential.The effectiveness and performance of the module are verified by experiments on different platforms using different coding methods.Afterwards,aiming at the data transmission in the weak network environment,this paper designs the transmission module of surveillance video.Concurrent transmission are used to speed up the transmission speed of data in the network in this paper.At the same time,the context information of the connection is maintained on the server side to avoid repeated data transmission.The experiments on public network and local network prove the effectiveness of the data transmission method in this paper.Finally,this paper designs a content analysis module.Based on machine learning technology,support vector machine classifier is used as the target detector.The target detecting unit with environmental adaptability is obtained by training the classifier with input samples in advance.The surveillance videos captured by the system are the data source of the content analysis module and the target detection results are saved in the database.We also design a user interface system for the user.The video content and the analysis result could be browsed in this system,which improves the efficiency of user analysis.At last,we design experiments to compare the result of the method in this paper with the result of the existing method,proving the improvement of this paper.