1 / 13

Kinect Image Manipulation Project

Kinect Image Manipulation Project. CSC410 Victor Milenkovic , Justin Stoecker , Brendan Loyot. Project Overview. Build a system to manipulate 3d images on a screen with the Kinect device. 1 st steps are to build up a set of gestures that the K inect recognizes.

lamis
Download Presentation

Kinect Image Manipulation Project

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. Kinect Image Manipulation Project CSC410 Victor Milenkovic, Justin Stoecker, Brendan Loyot

  2. Project Overview • Build a system to manipulate 3d images on a screen with the Kinect device. • 1st steps are to build up a set of gestures that the Kinect recognizes. • Key input in form of a depth map. • Recognize gestures by using statistical and physics methods. Key properties: Depth map, histogram, standard deviation, principal moments and axes of inertia.

  3. Depth Map • Values processed by the Kinect and received by this system. • 2D array. The X,Y axes match the X,Y positions of a given pixel and the value and X,Y is the Z position (i.e. depth). • From this we have a 3d representation of a pixel.

  4. Sample Depth Map

  5. Histogram • A histogram is a graph of the frequency with which a value occurs in a set. For our goals, we will create a histogram mapping the frequency of Z values. • We expect to see a spike in the histogram at 2 points: the Z range of the hand and the Z range of the wall. • We can isolate those pixels which belong to the hand by looking at this first spike.

  6. Sample Histogram

  7. Standard Deviation • The standard deviation represents how “spread out” a distribution of points is. • We can use this to determine whether the hand is open or closed, for example. With the hand open, we expect a larger standard deviation, and with the hand closed, we expect smaller.

  8. Inertia Tensor • If you don’t have given axes of rotation, you can create a generalized 2nd order tensor for the scalar moment of inertia. • For the Inertia Tensor when i=j, represents the moment of inertia around the X,Y,Z axes when the object is rotated around the X,Y,Z axes. For the other 6 terms, represents the moment of inertia around the ith axis when the object is rotated around the jth axis.

  9. Principal Moments and Directions of Inertia • Spectral Theorem says that there exists a coordinate system where • Essentially, this coordinate system is balanced around the symmetry of the object. • In this matrix, are the principal moments of inertia, and the axes with which this coordinate system was built are the principal axes.

  10. Eigenvalues and Eigenvectors • The process of finding this diagonalized matrix I and the axes of inertia involves a linear algebra technique for finding Eigenvalues and Eigenvectors. • Solving yields 3 values for

  11. Eigen values and vectors cont. • Solving Where is one of the 3 given’s above yields 3 Eigenvectors. The Eigenvalues correspond to the principal moments of inertia and the eigenvectors correspond to the principal axes.

  12. Importance of these inertial values • Once we have the moments of inertia, we can infer several valuable pieces of information. • For example, information about the rotation of the hand can be found from the angular momentum • The kinetic energy can be used to find the magnitude of movements of the hand.

  13. Current output

More Related