Fast pedestrian detection using LBP-based patterns of oriented edges

概要

論文の詳細を見る
This paper introduces a simple yet efficient algorithm for pedestrian detection on low resolution images. The main objective is to create a successful means of achieving a real-time pedestrian detection application. While the framework of the system consists of edge orientations combined with the LBP feature extractor, a novel way of selecting the threshold is introduced. With the objective being an efficient car vision algorithm, it is assumed that the negative samples in this context are mainly uniformly textured objects such as trees, roads, and buildings. This threshold improves significantly the detection rate as well as the processing time. Furthermore, it makes the system robust to uniformly cluttered backgrounds, noise and light variations. The test data is the INRIA pedestrian dataset and for the classification, a support vector machine with an RBF kernel is used. The kernel parameters are chosen to maximize the AUC of the receiver operating curve (ROC). The system performs at a state-of-the-art detection rates while being intuitive as well as very fast which leaves sufficient processing time for further operations such as tracking and danger estimation.
2012-03-22