Segmentation approach using local image statistics

被引：7

作者：

Mendonça, AP

da Silva, EAB

机构：

[1] COPPE, IME, BR-2290270 Rio De Janeiro, Brazil

[2] UFRJ, EE,DEL, COPPE, PEE, BR-21945970 Rio De Janeiro, Brazil

来源：

ELECTRONICS LETTERS | 2000年 / 36卷 / 14期

关键词：

Algorithms - Image analysis - Image quality - Statistics - Vectors;

D O I：

10.1049/el:20000869

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

An efficient segmentation technique is described that uses local application of the K-means method with feature vectors based on image statistics. it requires little computational effort and works well for different classes of images, comparing favourably ro others in the literature.

引用

页码：1199 / 1201

页数：3