SMARTEYE for human posture identification

Abstract

This report provides the summary of my works in my Final Year Project – SMARTEYE for human posture identification. A project aims to create an intelligent visual surveillance system to monitor safety of the elderly or infirm at home. For that objective, it is crucial to develop a human detection technique that can track and identify people from a live stream correctly and efficiently. In this project, I have implemented three common methods found in the literature with varying complexity and performance to find out the most effective methods for the goal. At the end of the project, all methods were implemented successfully, the methods implemented can be used to extracts the human figure from a video, or supplies the results to further post-processing tasks.