Research Of Vehicle Detection And Tracking Based On Monocular Vision

Intelligent Transportation System is applied to manage the entire traffic system, integrated advanced information technology, data communication transmission technology, electronic sensor technology, electronic control technology and computer processing technology effectively, creating a wide range, all-round role, real-time, accurate and efficient integrated transportation management system.The effective detection and real-time tracking of vehicle, plays an important role in the Intelligent Transportation Systems. There are many difficulties in the process of Vehicle detection and tracking, vehicle's own shadow, or blocked other vehicle, or obscured by background objects. For these difficulties, this dissertation presents real-time vehicle detection, tracking method. Major works are as follows:1. In order to solve the impact of the shadow and the highlighting etc. in the process of vehicle tracking under fixed camera, we present an improved codebook model detection algorithm. This method deal with vehicles sequences directly in the YUV Color Space, sampled background values are quantized into codebooks, input pixel values of new frame are compared with the codebooks to identifying foreground areas, using a specific color judgments model, to determine the shadow or highlighted area.2. In the process of vehicle tracking, vehicles blocking each other and vehicles are obscured by background objects is very common, response to this problem, this article uses Kalman-based prediction method to track vehicle. Our tracking method using a prediction based on Kalman model and matching vehicle based area, predict the state of vehicle and match vehicle effectively. This method reduces not only the search scope but also the possibility of false matches.Experiments show that this algorithm can detect moving objects in complex traffic scenes effectively and rapidly, the proposed method can handle shadows, highlights, occlusion and the change of background, is efficient in computation, can meet the needs of real-time tracking.