摘要
提出一种综合图像的色彩信息、灰度信息和空间信息提取图像的特征向量的方法。该方法首先将图像从RGB空间转化为HSV空间，并进行非均匀量化；其次将彩色转化为灰度图像；然后将图像划分成互有重叠的4块，分别求出每块图像的色彩直方图、灰度直方图；最后采用这8个直方图作为图像的特征向量对图像进行检索。
This paper presents a method of using the information of color, grey level and space as image feature. In this method, images are converted from RGB space to HSV space and colors are divided unequally , then images are divided into 4 part overlaply and the color histogram and grey level histogram are extracted. At last, these histograms are used as image features to retrieval image.
出处
《计算机工程》
CAS
CSCD
北大核心
2004年第10期20-22,共3页
Computer Engineering
基金
辽宁省科技基金资助项目(2020821)
关键词
基于内容图像检索
图像数据库
图像特征
Content based image retrieval(CBIR)
Image database
Image features
