The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Online Program Home
Abstract Details
Activity Number:
|
647
|
Type:
|
Topic Contributed
|
Date/Time:
|
Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
|
Sponsor:
|
International Indian Statistical Association
|
Abstract - #305714 |
Title:
|
Spatially Weighted Principal Component Analysis for Image Classification
|
Author(s):
|
Ruixin Guo*+ and Hongtu Zhu
|
Companies:
|
University of Colorado and The University of North Carolina at Chapel Hill
|
Address:
|
, , , United States
|
Keywords:
|
PCA ;
Image classification
|
Abstract:
|
Two challenges of classification for image data are the high dimensionality of the feature space and the complex spatial structure on a two-dimensional (2D) surface or in a 3D volume. Thus, dimension reduction prior to classification becomes important and necessary for image data. However, commonly used dimension reduction methods do not take into account of the underlying spatial information possessed by image data. This work develops a Spatially Weighted Principal Component Analysis (SWPCA) method by introducing two sets of weights (global and local spatial weights) to the reconstruction error, which enables a selective treatment of individual features and has the ability of incorporating not only the class label information but also the local spatial pattern of image data. A two-step iterative algorithm is provided for SWPCA, along with proposed weight determination procedures. We apply our method to both simulation studies and a real data analysis. Our proposed SWPCA demonstrates substantial improvement over commonly used PCA methods based on the classification results.
|
The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.
Back to the full JSM 2012 program
|
2012 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.