We have investigated if it is possible to detect and classify three different vehicles in real time by use of a 'smart' IR camera. The vehicles were a light armored tank destroyer, an all terrain vehicle, and an oversnow vehicle. The smart IR camera is simulated using a commercially available smart camera for the visual wavelength band. The camera is focused on a monitor where we are replaying IR sequences on the vehicles, which were recorded in the field. Classification of the different vehicles is done on the basis of geometrical measurements in segmented images. On one hand, a plain geometrical classification is done and the criteria used are that the measurements of height, length, and area of the vehicle have to be in certain intervals. On the other hand, the classification is done using a neural network. The neural net gives the probability for the different kinds of vehicles. The results show that it is possible to detect and classify the vehicles in real time. A neural network considerably increases the possibility of correct classification.


    Access

    Access via TIB

    Check availability in my library


    Export, share and cite



    Title :

    Upptaeckt och Klassning av Fordon med Simulerad Smart IR-Kamera (Detection and Classification of Vehicles by Use of a Simulated Smart IR-Camera)


    Contributors:
    G. Carlsson (author) / T. Gustafsson (author) / U. Ringh (author)

    Publication date :

    1994


    Size :

    26 pages


    Type of media :

    Report


    Type of material :

    No indication


    Language :

    English




    Lastindikatorer för fordon

    Formgren, Carl | TIBKAT | 1973


    Sidostabilitet hos ramstyrda fordon

    Lindberg, Tord | TIBKAT | 1976



    Miljoemedveten farthaallning i fordon

    Palmquist, U. / Swedish Road and Traffic Research Institute | British Library Conference Proceedings | 1991