| 基于金字塔梯度方向图像特征的检索模型设计 |
| |
| 引用本文： | 白宗文,周美丽,白 茹. 基于金字塔梯度方向图像特征的检索模型设计[J]. 现代电子技术, 2014, 0(15): 65-67 |
| |
| 作者姓名： | 白宗文 周美丽 白 茹 |
| |
| 作者单位： | 延安大学,陕西延安716000 |
| |
| 基金项目： | 陕西省教育厅应用研究项目：图像修复技术在延安文物保护中的应用与研究（12JK1004）;延安市工业攻关项目：基于机器视觉的文物修复系统修复与实现（2012kg-07） |
| |
| 摘 要： | 随着互联网的普及和发展,电子商务网站数量急剧增长,迫切需要一个平台对在线销售商品进行标注以方便用户进行搜索。通过提取类别图像和测试图像的金字塔梯度方向直方图(PHOG)全部特征,然后计算两者之间的距离,测试图像与类别图像距离比较近的就属于同一类图像。利用Matlab语言开发出了能够实现这一检索分类模型。实验证明这一模型灵活性好,准确性高。
|
| 关 键 词： | 电子商务 图像检索 混淆矩阵 金字塔梯度方向直方图 |
Design of image retrieval model based on image features in pyramid histogram of oriented gradient |
| |
| BAI Zong-wen,ZHOU Mei-li,BAI Ru. Design of image retrieval model based on image features in pyramid histogram of oriented gradient[J]. Modern Electronic Technique, 2014, 0(15): 65-67 |
| |
| Authors: | BAI Zong-wen ZHOU Mei-li BAI Ru |
| |
| Affiliation: | (Yan＇an University, Yan＇an 716000, China) |
| |
| Abstract: | With the development and popularization of Internet,and rapid increase of e-commerce web sites,sales of com-modity on line needs a platform for the annotation urgently to facilitate users to search in e-commerce site. The distance between the testing image and category image is calculated according to the entire features extracted in category image and testing image PHOG. If the distance between the testing image and category image is in the range,they belong to the same class. The model that could fulfil the image retrieval classification was developed with Matlab language. Experiment results show that the model has high flexibility and accuracy. |
| |
| Keywords: | e-commerce image retrieval confusion matrix PHOG |
| 本文献已被 CNKI 维普 等数据库收录！ |
