On the convergence of fitting algorithms in computer vision

被引：20

作者：

Chernov, N. ^{[1
]}

机构：

[1] Univ Alabama Birmingham, Dept Math, Birmingham, AL 35294 USA

来源：

JOURNAL OF MATHEMATICAL IMAGING AND VISION | 2007年 / 27卷 / 03期

基金：

美国国家科学基金会;

关键词：

statistical methods; maximum likelihood; convergence rates; conic fitting;

D O I：

10.1007/s10851-006-0646-1

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We investigate several numerical schemes for estimating parameters in computer vision problems: HEIV, FNS, renormalization method, and others. We prove mathematically that these algorithms converge rapidly, provided the noise is small. In fact, in just 1-2 iterations they achieve maximum possible statistical accuracy. Our results are supported by a numerical experiment. We also discuss the performance of these algorithms when the noise increases and/or outliers are present.

引用

页码：231 / 239

页数：9

共 16 条

[1] Statistical efficiency of curve fitting algorithms [J].

Chernov, N ;

Lesort, C .

COMPUTATIONAL STATISTICS & DATA ANALYSIS, 2004, 47 (04) :713-728

[2] FNS, CFNS and HEIV: A unifying approach [J].

Chojnacki, W ;

Brooks, MJ ;

van den Hengel, A ;

Gawley, D .

JOURNAL OF MATHEMATICAL IMAGING AND VISION, 2005, 23 (02) :175-183

[3] From FNS to HEIV: A link between two vision parameter estimation methods [J].

Chojnacki, W ;

Brooks, MJ ;

van den Hengel, A ;

Gawley, D .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2004, 26 (02) :264-268

[4] Rationalising the renormalisation method of Kanatani [J].

Chojnacki, W ;

Brooks, MJ ;

Hengel, AVD .

JOURNAL OF MATHEMATICAL IMAGING AND VISION, 2001, 14 (01) :21-38

[5] For geometric inference from images, what kind of statistical model is necessary? [J].

Kanatani, Kenichi .

Systems and Computers in Japan, 2004, 35 (06) :1-9

[6] Further improving geometric fitting [J].

Kanatani, K .

FIFTH INTERNATIONAL CONFERENCE ON 3-D DIGITAL IMAGING AND MODELING, PROCEEDINGS, 2005, :2-13

[7] Uncertainty modeling and model selection for geometric inference [J].

Kanatani, K .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2004, 26 (10) :1307-1319

[8] Cramer-Rao lower bounds for curve fitting [J].

Kanatani, K .

GRAPHICAL MODELS AND IMAGE PROCESSING, 1998, 60 (02) :93-99

[9] STATISTICAL BIAS OF CONIC FITTING AND RENORMALIZATION [J].

KANATANI, K .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1994, 16 (03) :320-326

[10]

Kanatani K., 1996, STAT OPTIMIZATION GE

← 1 2 →