1 / 8

3D View Simulation Based on Face Tracking

This presentation explores a 3D view simulation utilizing face tracking technology, designed for the EE7700 course. Leveraging a multi-touch user interface, the simulation allows users to physically interact with a virtual environment through motion sensing technologies like Kinect and infrared sensors. Key methodologies include Haar Cascade Classifier and CAMShift to effectively track facial motion, enabling translations and zooming within a virtual 3D scene. The experimental results demonstrate realistic interaction, replicating real-world movements in a simulated context.

tracey
Télécharger la présentation

3D View Simulation Based on Face Tracking

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. 3D View Simulation Based on Face Tracking Final Presentation for EE7700 DVP Shenghua Wan and Kang zhang May, 2012

  2. Motivation • Multi-touch User Interface • Physical Motion • Virtual Environment • Motion Sensing Game Consoles • Wii • Kinect Infrared projector and camera Infrared sensor bar Infrared LED

  3. Objective • Face Motion • Tracking • Translations • Zoom • Virtual 3D Scene Explore • 3D scene e.g. a Cube with 8 points with different depth values. • View Simulation

  4. Methodology-Face Tracking 1 • Haar Cascade Classifier (Viola & Jones 2001) • Haar-like features • Integral Image • AdaBoost (Freund & Schapire 1995) • Cascading • Implementation • OpenCV • Trained Classifier • Some Untuned Parameters

  5. Methodology-Face Tracking 2 • CAMShift ( Continuously Adaptive Mean-Shift ) • Assumption Image histogram of foreground object is time-invariant. • Back Projection • Mean Shift Algorithm • Locate new search window and goto 2 • Implementation • OpenCV

  6. Methodology-Face Tracking 2(cont) • Some comments on CAMShift • Sensitive and fast • Not Robust • tend to be interfered by objects with similar color distribution. fingers face arm even notebook!

  7. Experimental Results • Human face moves in real-world • Viewpoint moves in the simulated 3D scene as if we are looking at the real-world objects.

  8. Thank you!

More Related