Software Design And Implementation Of Multi-target Image Tracking Module

The traditional artillery reconnaissance is manual reconnaissance. There are many deficiencies in this way: First, the limited capacity of staff to capture objectives, and with surveillance staff is small. When there are many targets, it must temporarily give up some targets, such targets can lose some valuable;Second, is the target mobile, the time is too short tokeep up the speed of artificial reconnaissance; Third, in many harsh environments, manually monitor the targets will be very dangerous and difficult, the viability of the scouts is a test; Fourth, targets through the manual to convey information, dependence on a high professional quality inspector; Fifth is dependent on communications technology can not be objective information in real-time communication to the higher authorities.In response to these problems, this article tracking for multi-objective research is the use of automatic target tracking the instrument automatically target surveillance and tracking, but also realize many of targets automatic tracking: First, limited personnel and time shorter case, by automatic tracking, higher artillery reconnaissance capabilities; Second, in adverse environment,it can be automatically placed in the monitoring area monitors, remote operation by staff to achieve targets of surveillance and tracking, reconnaissance personnel improve survival; Third, through the multi-targets monitor the operation of a surveillance area can be increased so that efficiency will increase; Fourth is automatic detection system after the integration with the automatic command system, commanders can observed target real-time from the monitor, and select key target detection by information processing system , at last combat command issued.This allows for complete automation of tracking targets.The main content of this article:1.The establishment of a multi-target image based on video tracking system.2.Proposed adaptive threshold method to determine threshold. The threshold method for determining the image of unexpected noise, or relatively large changes in background intensity, the whole images will not be possible with an appropriate split a single threshold, with good results. 3.Gaussian mixture difference method proposed target detection algorithm, the method less affected by light conditions, can accurately detect targeting, stability and strong anti-interference.4.Proposes to use statistical methods to match multiple targets to achieve the tracking of multiple targets.5.Target image tracking system for multi-application put forward ideas battlefield.