Template matching: Matched spatial filters and beyond

被引：59

作者：

Brunelli, R ^{[1
]}

Poggio, T ^{[1
]}

机构：

[1] MIT,ARTIFICIAL INTELLIGENCE LAB,CAMBRIDGE,MA 02139

来源：

PATTERN RECOGNITION | 1997年 / 30卷 / 05期

关键词：

template matching; correlation; neural networks; learning; hyperBF networks; principal components; SYNTHETIC DISCRIMINANT FUNCTIONS; CORRELATION-ENERGY FILTERS; NOISE; RECOGNITION; VARIANCE;

D O I：

10.1016/S0031-3203(96)00104-5

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Template matching by means of cross-correlation is common practice in pattern recognition in spite of its drawbacks. This paper reviews some results on how these shortcomings can be removed. Several techniques (Matched Spatial Filters, Synthetic Discriminant Functions, Principal Components Projections and Reconstruction Residuals) are reviewed and compared on a common task: locating eyes in a database of faces. New variants are also proposed and compared: least squares Discriminant Functions and the combined use of projections on eigenfunctions and the corresponding reconstruction residuals. Finally, approximation networks are introduced in an attempt to improve filter design by the introduction of nonlinearity. (C) 1997 Pattern Recognition Society.

引用

页码：751 / 768

页数：18

共 32 条

[1]

[Anonymous], PROC IJCAI

[2]

[Anonymous], 1993, Advances in neural information processing systems

[3]

BASRI R, 1989, RECOGNITION LINEAR C

[4] A Novel Approach for Template Matching by Nonorthogonal Image Expansion [J].