Co-occurrence map: Quantizing multidimensional texture histograms

被引：14

作者：

Oja, E

Valkealahti, K

机构：

[1] Helsinki University of Technology, Lab. of Comp. and Info. Science, FIN-02150 Espoo

来源：

PATTERN RECOGNITION LETTERS | 1996年 / 17卷 / 07期

关键词：

texture analysis; multidimensional histograms; self-organizing map;

D O I：

10.1016/0167-8655(96)00018-9

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Co-occurrence matrices, multidimensional co-occurrence histograms, and histograms reduced by vector quantization with the self-organizing map were compared in the classification of monochrome and color textures. Increasing the histogram dimensionality improved the classification. The highest accuracy was obtained with the reduced histograms.

引用

页码：723 / 730

页数：8

共 6 条

[1]

Kohonen T., 1995, SELF ORG MAPS

[2]

KOIKKALAINEN P, 1994, P ECAI 94 11 EUR C A, P211

[3] PERFORMANCE EVALUATION FOR 4 CLASSES OF TEXTURAL FEATURES [J].

OHANIAN, PP ;

DUBES, RC .

PATTERN RECOGNITION, 1992, 25 (08) :819-833

[4]

OJA E, 1995, P IEEE INT C NEUR NE, V2, P1160

[5]

Ojala T., 1994, P 12 IAPR INT C PATT, V1, P582, DOI DOI 10.1109/ICPR.1994.576366

[6] TEXTURE FEATURES FOR CLASSIFICATION OF ULTRASONIC LIVER IMAGES [J].

WU, CM ;

CHEN, YC ;

HSIEH, KS .

IEEE TRANSACTIONS ON MEDICAL IMAGING, 1992, 11 (02) :141-152

← 1 →