Multiple-Instance Learning Support Vector Machine Algorithm based Pedestrian Detection

Pedestrian detection is one of the significant task for the intelligent transportation system.so the pedestrian detection become much relevant in automotive field for improving the safety systems. Many of the existing research papers focused on efficient pedestrian detection. The challenges that are facing the existing researches are (1)Various style of clothing appearances (2)Different possible poses of pedestrian (3) Presence of hidden objects(4)Frequent occlusion. In this paper, to resolve the above challenges by formulate the pedestrian detection based on the multiple instance learning support vector machine algorithm (MILSVM).Therefore the pedestrian performing various action will be accurately detected.