Pedestrian Detection in Far-Infrared Daytime Images Using a Hierarchical Codebook of SURF

• Published in: Sensors (Basel, Switzerland) Vol. 15; no. 4; pp. 8570 - 8594
• Main Authors: Besbes, Bassem, Rogozan, Alexandrina, Rus, Adela-Maria, Bensrhair, Abdelaziz, Broggi, Alberto
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 13.04.2015; MDPI
• Subjects: Automobile safety; Cameras; Computer Science; Computer Vision and Pattern Recognition; Daytime; Detectors; Engineering Sciences; far-infrared images; hierarchical codebook; Image detection; Machine Learning; Modules; pedestrian classification and tracking; pedestrian detection; Pedestrians; scale-invariant feature matching; Sensors; Signal and Image processing; State of the art; Statistics; Support vector machines; SURF; SVM; Vehicles; Vision systems; France
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s150408570

One of the main challenges in intelligent vehicles concerns pedestrian detection for driving assistance. Recent experiments have showed that state-of-the-art descriptors provide better performances on the far-infrared (FIR) spectrum than on the visible one, even in daytime conditions, for pedestrian classification. In this paper, we propose a pedestrian detector with on-board FIR camera. Our main contribution is the exploitation of the specific characteristics of FIR images to design a fast, scale-invariant and robust pedestrian detector. Our system consists of three modules, each based on speeded-up robust feature (SURF) matching. The first module allows generating regions-of-interest (ROI), since in FIR images of the pedestrian shapes may vary in large scales, but heads appear usually as light regions. ROI are detected with a high recall rate with the hierarchical codebook of SURF features located in head regions. The second module consists of pedestrian full-body classification by using SVM. This module allows one to enhance the precision with low computational cost. In the third module, we combine the mean shift algorithm with inter-frame scale-invariant SURF feature tracking to enhance the robustness of our system. The experimental evaluation shows that our system outperforms, in the FIR domain, the state-of-the-art Haar-like Adaboost-cascade, histogram of oriented gradients (HOG)/linear SVM (linSVM) and MultiFtrpedestrian detectors, trained on the FIR images.