基于分块主色调和纹理特征的彩色图像检索

提出一种基于分块主色调和纹理特征的彩色图像检索方法。该方法首先提取占彩色图像比重较大的前几种主色调,根据主色调将图像划分为不同分块,并以主色调作为各分块图像的颜色特征。然后,采用灰度共生矩阵算法,提取各分块的纹理特征。综合考虑各分块图像的颜色特征和纹理特征,从图像数据库中检索相似图像。实验结果表明,该方法具有较高的检索效率。

Image Retrieval based on Main Color and Texture Feature of Each Block

A new image retrieval algorithm based on the main color and texture of image blocks was presented.Firstly,the main kinds of color of each image were extracted as the image's color feature.Then images were partitioned into sub-images.Secondly,the texture of each block was extracted by gray symbiotic matrix algorithm as the image's texture feature.Color feature and texture feature were considered to retrieve similar images.Experiment results show that the new algorithm is more effective in image retrieval.