The application of dual thresholds binarization algorithm for pedestrian segmentation in thermal imaging

There is evidence that the major safety problem caused by darkness is the pedestrian collisions. Thermal Imaging systems can further reduce this risk. The image binarization is a preprocessing task, it play a very important role in pedestrian detection system. This paper applied the dual thresholds algorithm to process the thermal images for pedestrian segmentation. By a large number of experiments, we get the optimal combinations of the parameters of the algorithm, and get the good image binarization results of the thermal imaging.