People detection and tracking in videos

• Main Author: Chen, Jiaying
• Other Authors: Yuan Junsong
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/71294
• Institution: Nanyang Technological University

People detection and tracking in videos have a wide variety of applications in computer vision such as surveillance, people recognition, crowd behavior analysis, human-machine interaction. In this paper, I present some methods for people detection and tracking. Firstly, compared to the original HOG-SVM classifier without hard examples, I used the HOG-SVM classifier with pre-trained hard examples for people detection. Secondly, I proposed a novel method for people detection and tracking. Proposed approach utilizes the deformable part model (DPM) object detector to get people features and detect people positions in the video as well as high speed tracking with kernelized correlation filter (KCF) based tractor to tract the detected person.