本文已被：浏览 696次 下载 565次
投稿时间：2010-07-12
投稿时间：2010-07-12
中文摘要: 提出了一种基于颜色特征的图像摘要算法.首先将原始图像的RGB颜色模型转换到HSI模型中进行量化并转换为一维矢量,然后将图像分为64块,取得每个块内颜色的特征均值,最后进行量化产生固定长度的图像哈希序列.实验结果表明,相似内容的图像哈希距离较小,而不同内容的图像哈希距离较大.该算法还具有一定的篡改检测能力.
Abstract:An image hashing method based on color features is proposed.Firstly,the RGB color model is modified to HIS model.Then the color features are quantified to one dimension vector.The image is split to 64 blocks and the mean color feature is obtained.Finally,the image hashing value is obtained by quantifiing the mean color feature.The experimental results show that the distance between similar images is small,while the distance of different image pair is big.At the same time,this method can detect the color tampering.
keywords: image hashing color feature hue saturation intensity
文章编号：20100519 中图分类号： 文献标志码：
基金项目:
引用文本：
赵琰,周俊杰.基于颜色特征的图像摘要算法[J].上海电力大学学报,2010,26(5):489-492.
ZHAO Yan,ZHOU Jun-jie.Image Hashing Method Based on Color Feature[J].Journal of Shanghai University of Electric Power,2010,26(5):489-492.
赵琰,周俊杰.基于颜色特征的图像摘要算法[J].上海电力大学学报,2010,26(5):489-492.
ZHAO Yan,ZHOU Jun-jie.Image Hashing Method Based on Color Feature[J].Journal of Shanghai University of Electric Power,2010,26(5):489-492.