Research And Application Of Vehicle Queue Length Detection Algorithm Based On Video And Image Processing

As people’s living standards improve, the car ownership is increased rapidly. The expansion of the number of private cars brings great convenience for people’s life, at the same time, the emerging of the traffic problem in intelligent traffic system (ITS) of the demand is higher and higher. Vehicle queue length detection based on video image processing technology is an important research in intelligent traffic management system content, the purpose of which is real-time accurate and effective access to road traffic in the lane information, to ease traffic congestion and so on have great practical significance. In this paper the intelligent traffic management system of vehicle queue length detection based on video technology done in-depth research, the concrete content is as follows:Video image preprocessing. Experimental study of several commonly used algorithm for image denoising and edge detection algorithm using Gaussian smoothing filter for image denoising, edge detection using the Canny operator of the lane image preprocessing, experimental verification of the performance of the algorithm, showing good results, to lay the foundation for subsequent vehicle detection.Camera calibration. According to the position of the camera and the attribute parameter of the camera, also the imaging model and the point on two-dimensional image mapping between the points in three dimensional space, to restore the three-dimensional information of the object, thereby to calculate the actual vehicle queue length, the process of determining the camera parameters and the mapping relationship is called camera calibration. This experiment propose a improved camera automatic calibration technology, saving manpower, while improving the accuracy of the allocative efficiency of the system when it is actually in use, and the sensing distance.Vehicle queue length detection. According to the characteristics of the actual vehicle traffic queuing situation and video images, the present study proposes a sliding window algorithm based on the virtual three-dimensional space to a two-dimensional image map can be automatically retractable, this algorithm can be more accurately locate the vehicle than the traditional algorithm, so that identification vehicle and calculations vehicle queue length. By combining the vehicle contour background subtraction and edge detection algorithms derived to determine whether the presence of the vehicle to determine whether the vehicle movement based on frame difference method based on the duty cycle of the virtual non-zero pixels in the sliding window, and thus determine whether to line up the tail to calculate the vehicle queue length, experimental results indicate the approach is simple and acts stably.Finally, in this paper, the algorithm of vehicle queue length detection based on the video has been carried on the simulation test.