New Vehicle System Identification Based On Magnetic Feature Map

Vehicle detection is one of the important solutions for transposition system problems and can be utilised for different applications. Vehicles consist of deferent iron parts that lead to having a magnetic behaviour or map. This paper presents a new vehicle detection and identification system using magnetic sensors. A specific new system has been designed consisting of a non-magnetic material frame. Four magnetic sensors to detect the magnetic fields around the vehicle have been fixed in the frame in suitable places. Each sensor signal has been measured and transmitted to a PC for storing, analysing and identifying. Signals from magnetic sensors detected when the vehicle passes through the frame. A new database has been built employing the designed identification system used as reference data or reference signals. Two approaches have been used to analyse and identify the vehicle type. The first approach uses normalised cross-correlation depending on Fast Fourier Transform to find the similarity between magnetic sensors signals and the reference signals. The second approach uses a matching pattern approach in which the signals from sensors converted to an image and make a matching between the two patterns to identify the vehicle. The accuracy of the first approach of vehicle identification reached 90%, and the accuracy for the second approach reached 95%. These two approaches, compared with other approaches, gave more accurate results to identify the type of vehicle and distinguish between vehicles of the same type and give a signature to each type/model of the vehicles. Other approaches just detect the vehicle and identify if it is a small or big vehicle.