摘要
图像检索是计算机视觉研究的热点问题之一。首先对数字图像进行边缘检测;接下来对各个边缘点求其梯度;然后根据各个边缘点的梯度求得边缘梯度方向直方图;最后用街区距离进行相似性度量,实现图像检索。仿真实验表明,本文提出的算法是一种有效的图像检索方法。
Image retrieval is the has become the hot spot of computer vision. Firstly, we edge detection of digital image. Secondly, eompute the gradient of the edge feature points. Thirdly, get the edge gradient orientation histogram base on the gradient of the feature points. Lastly, use city block distance to similarity measurement, implement of image retrieval. Simulation results show that, the proposed algorithm is an effective method for image retrieval.
出处
《科技视界》
2012年第20期24-25,共2页
Science & Technology Vision
基金
湖南省科技计划项目
课题编号:No.2011GK3172
关键词
边缘检测
图像检索
形状特征
边缘梯度方向直方图
Edge detection
Image retrieval
Shape feature
Edge gradient orientation histogram
