Research on vehicle detection and classification algorithm by TMR sensor

Vehicle detection and classification is one of the key techniques in intelligent transportation system, and it has a broad application prospect. This paper studies the working mechanism of tunneling magneto resistance (TMR) sensor, and proposes a TMR sensor based vehicle detection technique by combining the double point dynamic sampling mechanism and the weighted Euclidean distance method for knee slopes of vehicle waveform. The experiments of road vehicle detection show that the feature of this detection method is low power consumption and long service life comparing with the induction coil detection; and it is not easily affected by external environment. Moreover, it can detect the vehicle flow and speed accurately and complete the identification and classification of vehicle types.