This study aimed to classify vehicles according to their categories, consisting of motorcycles, light vehicles, and heavy vehicles. For this purpose, there were three main techniques discussed: vehicle detection using Background Subtraction, feature extraction using Binary Robust Invariant Scalable Keypoint (BRISK), and vehicle classification using the K-Nearest Neighbors (KNN) algorithm for most cases. The dataset consisted of432 images for the training stage and one video data for the testing stage. The system performance was evaluated by reviewing the BRISK threshold value ranging from 10 to 80 with a k-value on KNN of 6. Results showed that the highest F1 scores were 96%, 86%, and 67% for motorcycles, light vehicles, and heavy vehicles, consecutively.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    An Approach for Vehicle’s Classification Using BRISK Feature Extraction




    Publication date :

    2021-07-29


    Size :

    2415122 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Baltic's brisk business

    British Library Online Contents | 2006



    Vehicle's Model Classification Using a Vertical Stereo Camera

    Lee, C. H. / Ahn, J. P. / Kim, Y. M. | British Library Conference Proceedings | 2015


    Vehicle’s Model Classification Using a Vertical Stereo Camera

    Lee, Chi Hak ;Ahn, Jong Pil ;Kim, Young Mo | Trans Tech Publications | 2015


    Brisk Maneuvering Device for Undersea Vehicles

    P. R. Bandyopadhyay | NTIS | 1996