NONLINEAR GENERALIZATION OF POINT DISTRIBUTION MODELS USING POLYNOMIAL REGRESSION

被引：21
作者：
SOZOU, PD
COOTES, TF
TAYLOR, CJ
DIMAURO, EC
机构：
[1] Wolfson Image Analysis Unit, Department of Medical Biophysics, University of Manchester, Manchester
来源：
IMAGE AND VISION COMPUTING | 1995年 / 13卷 / 05期
基金：
英国工程与自然科学研究理事会;
关键词：
POINT DISTRIBUTION MODELS; POLYNOMIAL REGRESSION;
D O I：
10.1016/0262-8856(95)99732-G
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
We have previously described how to model shape variability by means of point distribution models (PDM) in which there is a linear relationship between a set of shape parameters and the positions of points on the shape. This linear formulation can fail for shapes which articulate or bend. We show examples of such failure for both real and synthetic classes of shape. A new, more general formulation for PDMs, based on polynomial regression, is presented. The resulting polynomial regression PDMs (PRPDM) perform well on the data for which the linear method failed.
引用
收藏
页码：451 / 457
页数：7
相关论文
共 20 条
[1]  
AZZALINI A, 1993, J ROY STAT SOC B MET, V55, P549
[2]  
BARBOUR AD, 1991, J ROY STAT SOC B MET, V55, P541
[3]  
CHARTERS GC, 1994, THESIS U MANCHESTER
[4]  
Cootes T. F., 1992, P BRIT MACH VIS C, P9, DOI DOI 10.1007/978-1-4471-3201-1_2
[5]   COMBINING POINT DISTRIBUTION MODELS WITH SHAPE MODELS BASED ON FINITE-ELEMENT ANALYSIS [J].
COOTES, TF ;
TAYLOR, CJ .
IMAGE AND VISION COMPUTING, 1995, 13 (05) :403-409
[6]   ACTIVE SHAPE MODELS - THEIR TRAINING AND APPLICATION [J].
COOTES, TF ;
TAYLOR, CJ ;
COOPER, DH ;
GRAHAM, J .
COMPUTER VISION AND IMAGE UNDERSTANDING, 1995, 61 (01) :38-59
[7]  
DIGGLE PJ, 1983, STATISTICAL ANAL SPA
[8]  
GRENANDIER U, 1991, HANDS PATTERN THEORE
[9]   MODEL-BASED IMAGE INTERPRETATION USING GENETIC ALGORITHMS [J].
HILL, A ;
TAYLOR, CJ .
IMAGE AND VISION COMPUTING, 1992, 10 (05) :295-300
[10]  
HINTON GE, 1992, ADV NEURAL PROCESSIN, V4
12