A Review on Multiple Vehicle Detection and Tracking in Dynamic Environment

In this paper we review a novel algorithm Supervised Learning which is AI based approach and giving the system to training using Haar training algorithm and Clustering algorithm for tracking multiple vehicle objects in dynamic environments wherein under illumination conditions and the surrounding infrastructure is known. The proposed technique relies on trained and tested data provided by a haar training algorithm from processing features in each frame and simultaneously detecting background and foreground vehicles to ensure robust detection without collision. The method AdaBoost classifier is to classify its different features based on appearance and its motion based extraction from the region of interest. Unlike other existing methods that track rigid objects using rigid representations and also we reviewed object tracking used an enhanced particle filter-based method for multiple vehicle tracking based on appearance-based free-form obstacle representations. During this process, the particle state is described by two components, i.e., the object’s dynamic parameters and its estimated geometry. In order to solve these issues, an extended kalman particle filter is used. By accurately modeling the multiple object geometry using the rectangular box instead of a 3-D box and at the same time, separating the position and speed tracking from the geometry at the estimator level, the proposed method combines the efficiency of the robust detection and handles occlusion and collision in a dynamic environment.