A Method for Background Establishment Using Accumulate Histogram for Detection of Object Trajectory for Video Tracking Applications

Abstract

There are a lot of challenges facing the process of tracking moving targets, particularly whenusing fixed or moving cameras, so there is a need to find appropriate solutions to solve the kindof problems. The process of tracking moving objects becomes easier when the background isfixed (i.e., static) in order to distinguish the moving objects from the static parts by differentoperations.In this research, we attempt to utilize a technique for building a background template for thisscene, and this model replaces when time changes are caused by weather adversity or dynamicillumination.We use the accumulative histogram for each block of the video frames to compute the backgroundvalue of that block. Hence, the most probable mean of the block will be adopted as the backgroundmean value, and this value can be subtracted for the values of pixels which belong to that block.The mean values of all corresponding blocks which lay in the taken successive block are usedto get the most redundant mean which is adopted as the background value of that block.For the purpose of identifying the moving objects and separating it from other objects as wellas the background we apply the method of region grown. Some features like ( the center of anyobject) are extracted, and they are used to fix the trajectory of the identified object by trackingthe changes in position coordinated from frame set to other set by using Euclidean distancemetric.