| Video surveillance system has been widely deployed in indoor buildings in China.With the increasing demand for indoor positioning,the traditional positioning method based on wireless communication with portable devices can not meet the current demand.With the gradual development of visual positioning technology and visual mapping technology,there is a solution of positioning using machine vision.Studying how to apply visual positioning and mapping technology to the personnel positioning of the monitoring system is of great help to improve the intelligent monitoring system and promote the informatization and intellectualization of the monitoring system.Firstly,this paper investigates various indoor positioning technologies at present.There are two problems in the current indoor positioning technology:first,the traditional positioning system based on wireless communication is vulnerable to the influence of signal strength,and its performance in accuracy is not very stable;Second,the traditional indoor positioning technology has high requirements for equipment installation,requiring intensive indoor deployment of signal transmitting devices,and the use cost and deployment cost are very high.The above shortcomings restrict the development and application of indoor positioning system.However,at present,the video surveillance system has achieved full market-oriented development in China.Perfect video surveillance systems have been deployed in most office buildings,schools and other public places.With sufficient advantages in deployment cost and coverage,the demand for location technology based on video surveillance has become a trend.Combined with computer vision,scanning the internal terrain of buildings and identifying and calculating the personnel position in monitoring,adding the function of indoor personnel positioning on the basis of the existing monitoring system is a valuable application in the scene where video monitoring is widely deployed.Facing the needs and shortcomings,this paper proposes a surveillance video personnel tracking and positioning system based on visual slam,including a three-dimensional reconstruction algorithm based on visual slam for scanning the indoor terrain of buildings,an external parameter calibration algorithm for measuring the position relationship between topographic map and surveillance camera,and a personnel tracking and positioning algorithm for identifying pedestrians and giving position information,Finally,based on the above results,the monitoring personnel positioning and tracking system is designed,and the functions of each part are tested.Firstly,this paper analyzes the requirements of the system,and then designs and implements a surveillance video personnel tracking and positioning system based on visual slam.Through a series of system tests,it is verified that the system can meet the use requirements of the scene and meet the design expectations. |