摘要
该文通过对图像进行分块,提取各子块颜色直方图构建出各图像的主颜色信息,建立语义矩阵,完成了图像的提取,通过相关反馈完成了图像的检索。实验表明,加入语义的图像检索速度显著提高,有效的缩减了高层语义和低层特征的鸿沟。
Through segmenting the images into pieces,extracting the histograms of every piece,dominant colors have been extracted.At the same time,through semantic matrix an d relevance feedback,image semantic extract and retrieval have been achieved.The result shows that the retrieval based on semantic has a good performance,and it shortens the gap between the high-level semantic and the low-level feature.
作者
车志强
魏迎梅
刘飞裔
CHE Zhi-qiang,WEI Ying-mei,LIU Fei-yi(National University of Defense Technology,Changsha 410073,China)
出处
《电脑知识与技术》
2008年第12X期2722-2724,共3页
Computer Knowledge and Technology
关键词
图像检索
主颜色
语义矩阵
相关反馈
image retrieval
dominant colors
semantic matrix
relevance feedback
