Accuracy Improvement Of Video-Based Vehicle Velocity Measurement

Video-based vehicle velocity measurement is a new technology for vehicle velocity measurement. The state-of-the-art of video-based vehicle velocity measurement is reviewed and factors influencing velocity measurement accuracy are analyzed, then a method to improve the velocity measurement accuracy in intelligent bayonet system is designed, and experiments are carried out. The main work of this paper is as follows:1. The mapping relationship between the image plane and the road coordinate is established by the properties of vanishing points in the video-based vehicle velocity measurement system. The coordinate of vanishing points in the screen is calculated by the principle of the perspective projection of the image. The mapping formula is determined according to the relationship among the vanishing points and the pitch angle of the road surface, declination, rotation of the camera’s internal angel, the pixel focal length, the distance of the two coordinates.2. The feature block of vehicle is precisely located by rough location and refined location to the external boundary of the second character in the license plate. The area of the second character is determined with the illumination compensation of license plate image, binarization, tilt correction, eliminate noise, segmentation. Template matching is completed with wavelet analysis and the structural characteristic of the character. Then the boundary of the character is adjusted with the information of template.3. The vehicle velocity measurement is realized by the location of the vehicle feature block in the image sequence and the depth information of the feature block. Finally, the suggested scheme is tested in the intelligent bayonet system. Experimental results show that the proposed method can effectively improve the velocity measurement accuracy.