摘要
设计开发了一套基于角度结构描述符的彩色图像检索系统。角度结构描述符利用角度结构一种特征描述图像颜色、纹理、形状和空间布局信息等。它在HSV颜色空间根据图像块相邻像素之间的内部相关性提取方向的多样性,形成图像特征向量。利用MATLAB语言搭建基于角度结构描述符的彩色图像检索系统,系统具有图像特征提取、查询图像选取、图像相似性计算及优化、图像检索结果显示及性能评价等功能。在公开的Corel-1k图像库对系统进行图像检索性能测试,结果表明该系统可以有效检索出相似的图像。
A color image retrieval system based on angle structure descriptor was designed and developed.The angle structure descriptor can capture image color,texture,shape,and spatial layout information using angle features.It forms an image feature vector in the HSV color space based on the diversity of the internal correlation between the adjacent pixels of the image block.The MATLAB language was used to build the color image retrieval system based on angle structure descriptor.The system has functions of image feature extraction,query image selection,image similarity calculation and optimization,image retrieval result display and performance evaluation.The image retrieval performance of the system was tested on the public Corel-1 k image database,and the retrieval result showed that the system can retrieve similar images effectively.
作者
徐国清
谭景之
XU Guo-qing;TAN Jing-zhi(College of Computer and information Engineering, Nanyang Institute of Technology, Nanyang 473004, China)
出处
《南阳理工学院学报》
2018年第6期55-60,共6页
Journal of Nanyang Institute of Technology
基金
河南省教育厅高等学校重点科研项目(19A520029
17A520046
19B520017)
关键词
图像检索
角度结构描述符
特征向量
图像相似性
image retrieval
angle structure descriptors
feature vectors
image similarity
