Detekce objektů pomocí HOG, SVM a Random Forests

Abstract

This thesis is focused on pedestrian detection and licence plate recognition in images. Detectors are based on the histograms of oriented gradients (HOG). The final feature vector of HOG is then used as an input for a trainable classifier. In this thesis, the support vector machine (SVM) and random forests (RF) are used. The main ideas, experiments, and results are shown in this work.