Research And Implementation Of Shadow Removal And Traffic Parameter Extraction In Video Image

In recent years, laying new pavement or adding more lanes is becoming less and less feasible, thus that is no longer efficient solution for serious traffic congestion problem due to consistent increment of vehicles. Road traffic monitoring and control is the essential component of this solution. To monitor road traffic, it is necessary to extract traffic parameters that describe the characteristics of vehicles and their movement on the road. Vehicle counts, vehicle speed, vehicle path, flow rates, vehicle density are all example of useful traffic parameters. Various kinds of traffic control systems, for example, automatic tolls, congestion and incident detection, and increasing road capacity via automatic routing and variable speed limit, can be implemented with these traffic parameters.In this paper, we describe a method for multiple vehicles tracking and traffic parameters extraction based on Discrete Wavelet Transform. In order to describe the traffic characteristics, some important traffic parameters are extracted, such as vehicle trajectory, vehicle speed and vehicle count. Because the moving vehicles belong to low frequency part and we don't take more attention to the detailed information of the frame, so our background subtraction is performed in the low frequency sub-image of the two-scale DWT, and the disturbance of fake motions is effectively prevented.A shadow removal algorithm based on shadow attributes is presented in this paper. The direction of shadow is estimated first, and then shadow points are sampled based on the direction of shadow. At last the shadow attributes is calculated using the sampled shadow points. The shadow is removed based on shadow attributes. The vehicle detection error due to occlusion makes serious difficulties to extract traffic parameters. In our method the occlusion is effectively resolved.The CamShift algorithm is used to tracking object in this paper. The experimental results show that the CamShift algorithm is effective for face and vehicle tracking, but is semiautomatic in our experiment. How to improve this algorithm to automatic is our target.We simulate the algorithm in the computer and test the performance of traffic detection, shadow removal, and occlusion removal. Experimental results show that the algorithm is feasible and robust.