| 基于FSIM的彩色图像检索 |
| |
| 引用本文： | 马自萍,康宝生,马金林.基于FSIM的彩色图像检索[J].宁夏大学学报(自然科学版),2011,32(4):321-324. |
| |
| 作者姓名： | 马自萍 康宝生 马金林 |
| |
| 作者单位： | 1. 北方民族大学信息与计算科学学院,宁夏银川750021;西北大学信息科学与技术学院,陕西西安710120
2. 西北大学信息科学与技术学院,陕西西安,710120 |
| |
| 基金项目： | 国家自然科学基金资助项目(60873095,61001156);北方民族大学科学研究资助项目(2011ZMY004) |
| |
| 摘 要： | 为了得到一种简单而有效的图像特征，提出了一种频域结构相似度的图像检索方法．该方法利用彩色图像的高斯变换下相位不变、梯度的结构特征和局部二值模式描述图像特征，既弥补了相位不变性对图像对比度不变的影响，又能充分利用LBP纹理特征的旋转不变性，对彩色图像有很好的检索效果．为了验证该算法的效率，在3个彩色图像库中进行了对比实验．结果表明：该方法明显优于其他算法，比GCD—HSV，HSV，LBP算法的平均检索率分别提高了27．24％，29．86％，33．48％．
|
| 关 键 词： | 相位不变性 局部二值模式 梯度幅值 图像结构相似度 图像检索 |
Color Image Retrieval Based on FSIM |
| |
| Ma Ziping
,
Kang Baosheng
,
Ma Jinlin.Color Image Retrieval Based on FSIM[J].Journal of Ningxia University(Natural Science Edition),2011,32(4):321-324. |
| |
| Authors: | Ma Ziping Kang Baosheng Ma Jinlin |
| |
| Institution: | 1,2(1.School of Information and Computer Science,Beifang University of Nationalities,Yinchuan 750021,China; 2.School of Information and Science Technology,North University,Xi’an 710120,China) |
| |
| Abstract: | In order to obtain an effective and compound feature in image retrieval, a new algorithm is proposed based on frequency structural similarity （FSIM）. This method aims to extract phase congruent （PC）, gradient magnitude of an image smoothed with Gaussian transform, and local binary pattern feature （LBP）, which not only can fill up an efficiency of the contrast invariant of phase congruent, but also can fully utilize the rotation invariant of local binary pattern. In order to test the efficiency of the proposed method, the experiments are conducted in three databases. The experimental results show that the average retrieval precision of proposed method is respectively higher 27.24%, 29.86%, 33.48% than GCD_HSV, HSV, LBP. |
| |
| Keywords: | phase congruent local binary pattern gradient magnitude frequency structural similarity image retrieval |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
