Real-time infrared pedestrian detection via sparse representation

This paper presents a simple, novel, yet very powerful approach for real-time infrared pedestrian detection based on random projection. In our framework, firstly, a feature-centric efficient sliding window scheme is proposed for candidate pedestrians searching. Different from the traditional threshold or edge based region of interest (ROI) generation techniques, it performs robustly under different scenes without delicate parameter tuning. Secondly, at the feature extraction stage, a small set of random features is extracted from local image patches. To the best of our knowledge, this paper is the first to investigate random projection (RP) for infrared pedestrian feature representation. Finally, the random features in a pyramid grid are concatenated to perform sub-image classification using a support vector machine (SVM) classifier. In our case, both learning and classification are carried out in a compressed domain. Experimental results in various scenarios demonstrate the robustness and effectiveness of our method.