Application of cascades of classifiers in the vehicle detection scenario for the ‘SM4Public’ system

Abstract

In the paper, the use of cascading approaches for vehicle classification in static images is described. The problem concerns the selection of algorithms to be implemented in the ‘SM4Public’ security system for public spaces and is focused on specific system working scenario: the detection of vehicles in static images. Three feature extractors were experimentally evaluated using a cascading classification approach based on AdaBoost. The algorithms selected for feature extraction are Histogram of Oriented Gradients, Local Binary Patterns and Haar-like features. The paper contains brief introduction to the system characteristics, the description of the employed algorithms and the presentation of the experimental results.