Robust Vehicle Detection Algorithm

Abstract

The objective of this work is to develop an algorithm for the detection of vehicles in parking lot images. By the proposed method, each image region corresponding to a parking cell is segmented for initialization and temporal differencing is used for state tracking. Therefore, reference images taken in vacant state are not needed. The algorithm can easily be applied to parking lots in continuous service for the detection of the vehicles or to any application requiring the detection of the cars. Also the usage of the temporal differencing makes the system work faster.