A highway vehicle detection method based on the improved visual background extractor

When using the unmanned aerial vehicle (UAV) to detect highway vehicles, the movement of the UAV itself will cause great background changes, which in results will reduce the detection performance of the visual background extractor (ViBe). For this reason, this paper proposes a new vehicle detection method based on improved ViBe. First of all, the ViBe background model has been improved to achieve a better result of background segmentation. Then this paper uses several morphological image processing methods to process the binary image after the background segmentation, and then achieve the final vehicle detection. The experiment results show that this method can improve the accuracy of highway vehicle detection greatly.