Research on Video-based Moving Vehicle Detection Technology

Abstract

Moving vehicle detection technology has been used more and more widely in monitoring traffic conditions such as vehicle flow, vehicle speed, traffic accidents, and escape from violations. The most commonly used methods of moving vehicle detection technology include the inter-frame difference method, the background difference method, and the optical flow method. The inter-frame difference method subtracts the images of two or three adjacent frames, the pixel whose brightness change does not exceed the threshold value is considered as the background, and the pixel whose brightness change exceeds the threshold value is considered as the foreground; The background difference method uses a difference comparison between the current image and the background model to realize the moving vehicle detection; The optical flow method is to construct an optical flow field by changing the brightness value of each pixel point in an adjacent frame in a video image, and realize the detection of a moving vehicle through a moving field of the moving vehicle. The frame difference method can detect the contour edge information of a moving vehicle, and the detection result is less noisy; the background difference method can detect a more complete moving vehicle, including not only the contour information of the moving vehicle, but also the connected area inside the moving vehicle; The optical flow method can detect moving vehicles, but can only obtain the contour edge information of moving vehicles, and there are large voids in the interior of moving vehicles. In short, these methods can realize the detection of moving vehicles, but there are still some problems. With the research and implementation of improved algorithms, the detection technology of moving vehicles will be more and more widely used.