Image based Wheel Detection using Random Forest Classification
The aim of this master thesis is to detect and recognise wheels in images by means of image analysis. This could later on serve as a foundation for a safer vehicle counting and classification method than those currently in use that requires personnel to cross the lanes on installation. The general layout of the classification system consists of five stages: multi-scale transformation, window ext
