Journal of Emerging Trends in Engineering and Applied Sciences: Research Areas

Article Title: Development of Pedestrian Tracking System Using Lucas Kanade Technique
by Kazi Mowdud Ahmed, Firoza Naznin, Md Shahinuzzaman and Md Zahidul Islam

Abstract:
Pedestrian tracking is a rapidly evolving topic in many computer vision applications. This proposed system presents a real-time system for pedestrian tracking in sequences of grayscale images acquired by a stationary camera. The objective of this project is to track pedestrian object in successive video frames. A pedestrian moves in any direction in the road path. Our purpose is to track the single pedestrian see where he/she moves during a certain period of interest. Processing of this project is done at three levels: Detection of human for tracking, identifying feature of interest in image, tracking human through successive image frame. Object tracking is achieved through a combination of two methods: Lucas-Kanade features and connected components. This application performs tracking in real time, with previously learned training data. This project was implemented using OpenCV 2.0.

Keywords: pedestrian detection, codebook segmentation, optic flow, lucas kanade, tracking.

Download full paper