Abstract
Pedestrian detection is an increasingly interest research in computer vision with the challenging problem under complex background and occluded appearance in real world environment. The existing datasets have limitations for a large variation in human pose and clothing, variation of appearance, and cluttered backgrounds. In this paper, we considered the limitation of the existing dataset problem by providing with complex pose and occluded pedestrians from different views and complex backgrounds. Therefore, the main objective of this paper is to propose PSU Pedestrian Dataset for the Asian pedestrian environment which is different with the standard European datasets. For the performance comparison, PSU dataset and INRIA dataset are used to test with baseline Histogram of Oriented Gradient (HOG) and Support Vector Machine (SVM) learning model.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
