Contact me at 
| Keep in Touch | · | How people use Firefox | · | Travel Pets | · | Health care Interactive Narrative | · | MeetinBytes | · | FuturTech Conference | · | Nightmarket.web |
| Archive System Evaluation | · | Gesture Recognition | · |
Design process The project used OpenCV library to realize computer vision. First, I used some functions to find the area of palm in front of a simple background by statistically analyzing the contour of hand. The lower green dot locates the wrist.
Next, by applying GMM (Gaussian Mixture Models), fingers of user can be differentiated from the background.
The third step is to defined gestures. In this system, I implemented three different gestures: point with fingers, feast, and palm.
To demonstrate how the gestures work, I designed and programed two interfaces: one is similar to our PC desktop which allowing users to use gesture rather than mouse to control the object on the interface, and another is a weather map which also allow users to use the defined three gestures to move around whether icons on the map. {watch the demo on YouTube}
Mouse and keyboard are the dominant input hardware between human and machine at this time. Under this condition, our movement of control is not natural and many limitations occur during the operating process. Sometimes, our body even suffers chronic illness due to the constant operation. Therefore, this project aimed at developing a system which allows users to control with their hands by analyzing the input image information from a web cam.
Project DetailsMy role: This is an independent work for my undergraduate thesis. I designed the interface and built the system with openCV library. I also conducted usability testing with 36 different users to test my system
Duration: 10.2008 - 06.2009
Technology used: Processing (JavaScript), OpenCV library, C++, Photoshop, Illustrator