1 / 13

Statistical Shape Analysis

Statistical Shape Analysis. Research Institute of Computer Science and Technology, Ningbo University Zhu Shaojun. What is a shape?. Shape is all the geometrical information that remains when location, scale and rotational effects are filtered out from an object. Landmark.

cleary
Download Presentation

Statistical Shape Analysis

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. Statistical Shape Analysis Research Institute of Computer Science and Technology, Ningbo University Zhu Shaojun

  2. What is a shape? • Shape is all the geometrical information that remains when location, scale and rotational effects are filtered out from an object.

  3. Landmark • A landmark is a point of correspondence on each object that matches between and within populations. • Anatomical landmarks • Mathematical landmarks • Pseudo-landmarks

  4. Landmark • A mathematical representation of an n-point shape in k dimensions could be to concatenate each dimension into a kn-vector. The vector representation for planar shapes (i.e. k = 2) would then be:

  5. Shape Alignment • To obtain a true shape representation – according to our definition – location, scale and rotational effects need to be filtered out.

  6. The Procrustes Shape Distance • Compute the centroid of each shape. • Re-scale each shape to have equal size. • Align w.r.t. position the two shapes at their centroids. • Align w.r.t. orientation by rotation.

  7. shape size metric • A shape size metric S(x) is any positive real valued function of the shape vector that fulfils the following property (like all other metrics):

  8. shape size metric

  9. rotation • Arrange the size and position aligned x1 and x2 as n*k matrices (in the planar case k = 2). • Calculate the SVD, UDV(T), of x1(T)x2 in order to maximize the correlation between the two sets of landmarks. • The rotation matrix needed to optimally superimpose x1 upon x2 is then VU(T):

  10. Generalized Procrustes Analysis • Choose an initial estimate of the mean shape (e.g. the first shape in the set). • Align all the remaining shapes to the mean shape. • Re-calculate the estimate of the mean from the aligned shapes. • If the estimated mean has changed return to step 2.

  11. Modeling Shape Variation • PCA

More Related