基于直方图的图像信息检索方法  

Image Information Retrieval Method Based on Histogram

作　　者：毕军涛[1] 丁喜纲[1] BI Juntao;DING Xigang(Qingdao Vocational and Technical College of Hotel Management,Qingdao Shandong 266100,China)

机构地区：[1]青岛酒店管理职业技术学院,山东青岛266100

出　　处：《信息与电脑》2021年第4期46-47,共2页Information & Computer

摘　　要：传统检索方法不能有效利用图像之间紧密的相关交集,实用性较差。为此,本文提出基于直方图的图像信息检索方法,首先将原始图像转换为累加直方图,然后提取累加直方图的特征,并计算图像间的相似性度量,定义图像检索空间语言,最后建立检索模型,将直方图特征作为输入,结合检索语言和相似性度量实现图像信息检索。实验结果表明,与传统的检索方法相比,设计方法的实用性更强。Traditional retrieval methods cannot effectively use the close correlation and intersection between images,and their practicability is poor.To this end,this paper proposes an image information retrieval method based on histograms.First,the original image is converted into a cumulative histogram,then the characteristics of the cumulative histogram are extracted,the similarity measure between images is calculated,the image retrieval space language is defined,and the retrieval is finally established The model uses histogram features as input,and combines retrieval language and similarity measures to achieve image information retrieval.The experimental results show that the design method is more practical than the traditional retrieval method.

关 键 词：直方图 图像信息 相似性度量 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]