A hand detection algorithm from real-time imagery

Dutton, Zack. (2016). A hand detection algorithm from real-time imagery. [Final Year Project]

Microsoft Word Final Report - Zack Dutton.docx - Submitted Version Restricted to Registered users only Download (10MB)

Abstract

The basis of this project was to create a program capable of performing real-time hand detection from a given image from camera hardware. Various techniques and methods were applied to a given image to allow the program to detect a user’s hands and allow a form of interaction using sprites. The tools and resources included in the development of this project include the OpenCV software library that provides a number of useful image manipulation and image smoothing techniques used within this project. The Kinect 2.0 sensor and Kinect SDK were also used for this project to provide the different types of image given for this project. This project has explored several methods of how this can be approached and displayed the best result which provides a form of interaction with the user. Some of the techniques applied range from edge detection algorithms and image smoothing techniques all of which have been either explored or discussed within this project.

Actions (login required)

View Item