Pedestrian Detection for UAVs Using Cascade Classifiers with Meanshift

In this paper, we propose an algorithm for pedestrian detection focusing on UAV applications. Our proposal is based on a combination of Haar-LBP features with Adaboost for the training process, and Meanshift for improving the performance of the pedestrian detector. We mount a dataset with images captured from surveillance cameras. Our dataset and algorithm are evaluated and compared with other approaches from the literature.