Embedded image processing module for vehicle occupancy detection

Abstract

Due to the proliferation in the number of vehicles on the road, traffic problems are bound to exist as the current transportation infrastructure and car park facilities developed are not able to sustain the number of vehicles. With the intention of alleviating the problems which exist in the various car park facilities, a parking space occupancy detection module is proposed. The various vehicle detection sensors aplenty on the market have lead to the development of various vehicle detection systems with different functionalities. Herein, the research focuses on the usage of video cameras in occupancy detection within the car park facility. The images captured by the cameras will be processed to obtain the occupancy status of individual parking spaces. With the occupancy status constantly changing, the module developed must be able to handle recursive real – time computation of the occupancy status for all the parking spaces defined within the facility. As image processing is well known to be computationally intensive, the FPGA would be utilized in offering the additional benefits in terms of computational speed. The architecture proposed involves connection of multiple cameras to a central computer. It will also serve to relieve the burden of the central computer as it only need to receive the computation results from the FPGA device without having to perform heavy computing tasks.