Automatic detection system of the fallen down for the parked bicycle in the underground parking garage using optical flow method

Recently, there is a new parking device for the bicycles, which are built on the underground because of good landscape, space saving and more in the urban towns. However, there are a few accidents such that the bicycles parked with high speed by the powerful robot arm are sometimes fallen down in the underground garage after parking. This is serious problem for the use. So a real time detection system for the fallen down bicycles is requested, extremely. We propose a new and unique detection algorithm with applied optical flow method to solve this serious problem. We propose its detection mechanism and show its effectiveness using real monitoring image of the parked bicycle. Finally, we implement this detection algorithm to a Linux computer board and show its practical feasibility.