A method for estimating the speed of a vehicle moving in a single traffic lane is proposed. The blurring of the vehicle image and the presence of Gaussian noise at individual stages of image acquisition were taken into account. The algorithm is based on initial signal processing using Brown’s exponential smoothing method. The performance of the algorithm under conditions of strong random interference was verified through computer simulations and compared with other vehicle speed estimation methods.The paper addresses the issue of vehicle speed estimation based on an image recorded by a video camera. In the case of a low level of interference, a method consisting of subtracting successive image frames is used, which allows for the identification of moving objects. In the case of a high level of interference, the method of subtracting successive image sequences becomes useless because it leads to an increase in the noise level—the noise variance is summed. In this paper, where strong random interference is considered, it is assumed that the position of the object is estimated for individual frames, while the speed is determined based on sequences of images (10, 20, 30 frames). The work is a continuation of the topics discussed in the author’s previous works [2], [3]. (…)
The article contains 15624 characters.
Source: Logistyka Magazine