|Table of Contents|

Research on Sub-image Feature Fusion Rules in Trademark Image Retrieval(PDF)

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
2002年05期
Page:
508-511
Research Field:
Publishing date:
2002-10-30

Info

Title:
Research on Sub-image Feature Fusion Rules in Trademark Image Retrieval
Author(s):
SunXinghua GuoLi ① WangZhengqun ① YangJingyu ①
Department of Computer Science and Technology, Tsinghua University, Beijing 100084)
Keywords:
t rademark image ret rieval sub-image feature fusion PVR value
PACS:
TP391.4
DOI:
-
Abstract:
T his paper researches the sub- image feature fusion rules in the t rademark image retrieval, and improves the sub-image feature fusion rules by weig ht ing based on the sub- image reliability. Ex periments are performed using Hu invariant moments on the t rademark image ret rieval based on sub-image feature fusion, and the PVR value is used as the image retrieval performance evaluat ion criterion. Experiment s show that the trademark imag e ret rieval performance based on sub-image feature fusion outperforms the one based on global image features, and that the sub-image feature fusion w eighted rules outperform the unw eighted ones, in w hich the minimum weig hted average rule is optimal.

References:

1 庄越挺, 潘云鹤, 芮勇1 基于内容的图像检索综述. 模式识别与人工智能, 1999, 12( 2) : 170~177
2 Jain A K, Vailaya A. Shape-based retrieval: A case study w ith tr ademark image database. Patter n Recognition, 1998, 31( 9) : 1 369~ 1 390
3 徐旭1 基于视觉特征的图像检索系统研究: [ 博士学位论文] . 杭州: 浙江大学, 1999
4 姚玉荣, 章毓晋. 利用小波和矩进行基于形状的图像检索. 中国图像图形学报, 2000, 5( A3) : 206~ 210
5 金琪, 戴汝为. 矩、不变矩、角度矩及傅里叶描绘子. 模式识别与人工智能, 1993, 6( 3) : 181~188

Memo

Memo:
-
Last Update: 2002-10-30