摘要
针对图像检索中单一特征检索的不足,提出一种结合分块颜色特征和形状特征的图像检索方法,用空间局部信息增强全局颜色特征,采用分块颜色直方图提取图像的颜色特征,利用Hu不变矩提取图像的形状特征,以分块颜色和形状作为综合特征,采用相应的相似性度量方法进行图像匹配.实验结果证明了该方法的可行性,适用于基于区域或对象的图像检索.
Aiming at the deficiency of single feature retrieval in image retrieval, proposes an image retrieval method based on block color and shape features, global color features is enhanced by using the local spatial information, the color feature of image is extracted by using the block color histogram, shape feature of image is extracted by using Hu invariant moments,Integrated block color feature and shape feature, image matching using the corresponding similarity measure method. The experimental results prove the feasibility of the method, suitable for regional or object image retrieval.
出处
《泉州师范学院学报》
2013年第6期53-56,共4页
Journal of Quanzhou Normal University
基金
福建江夏学院青年科研基金资助项目(2012C027)
关键词
图像检索
颜色特征
形状特征
分块
image retrieval color feature shape feature block
