Search

Stereoscopic MRA acquires a pair of blood vessel projections at two different viewing angles. Previously, we have developed two algorithms to reconstruct 3-D blood vessels from stereoscopic MRA. The assumption we made was that blood vessels were tilting circular tubes and the shape of the vessel on every cross-section was an ellipse. Since an ellipse can be represented in either algebraic form or parametric form, our previous algorithms reconstructed the ellipses by representing them in these two forms. In this paper, we further improved the accuracy of our previous algorithms by an order through two enhancements. The first improvement we made was a better method to estimate the rotation angle of the major axis of an ellipse. Instead of using the center of two adjacent ellipses to estimate the rotation angle as in our previous method, the new method used the projection lengths of the two views to estimate the angle. The second improvement we made was the equation to describe the relationship between the major/minor axes and the projection lengths. In our experiments, the average estimation error for the parametric algorithm was improved from 0.471 pixels to 0.066 pixels. The average error for the algebraic algorithm was improved from 0.101 pixels to 0.014 pixels., (Copyright © 2011 Elsevier Ltd. All rights reserved.)

Abstract: Kernel-based nonlinear characteristic extraction and classification algorithms are popular new research directions in machine learning. In this paper, we propose an improved photometric stereo scheme based on improved kernel-independent component analysis method to reconstruct 3D human faces. Next, we fetch the information of 3D faces for facial face recognition. For reconstruction, we obtain the correct normal vector’s sequence to form the surface, and use a method for enforcing integrability to reconstruct 3D objects. We test our algorithm on a number of real images captured from the Yale Face Database B, and use three kinds of methods to fetch characteristic values. Those methods are called contour-based, circle-based, and feature-based methods. Then, a three-layer, feed-forward neural network trained by a back-propagation algorithm is used to realize a classifier. All the experimental results were compared to those of the existing human face reconstruction and recognition approaches tested on the same images. The experimental results demonstrate that the proposed improved kernel independent component analysis (IKICA) method is efficient in reconstruction and face recognition applications. [ABSTRACT FROM AUTHOR]