| 一种基于剪切波变换的人脸表情识别算法 |
| |
| 引用本文： | 穆晓敏,张众,曲艳,刘占卫. 一种基于剪切波变换的人脸表情识别算法[J]. 郑州大学学报(自然科学版), 2014, 0(3): 54-58 |
| |
| 作者姓名： | 穆晓敏 张众 曲艳 刘占卫 |
| |
| 作者单位： | 郑州大学信息工程学院,河南郑州450001 |
| |
| 基金项目： | 基金项目：国家自然科学基金资助项目,编号61331021. |
| |
| 摘 要： | 剪切波变换是一种新型多尺度几何分析方法.利用剪切波变换良好的方向敏感性和各向异性,提出了一种基于剪切波变换的人脸表情识别方法.实验表明,该方法在日本JAFFE和加拿大瑞尔森RML人脸表情数据库上的识别率分别达到了98.58％和95.83％.
|
| 关 键 词： | 剪切波变换 表情识别 支持向量机 |
A Facial Expression Recognition Approach Based on the Shearlet Transformation |
| |
| MU Xiao-min,ZHANG Zhong,QU Yan,LIU Zhan-wei. A Facial Expression Recognition Approach Based on the Shearlet Transformation[J]. Journal of Zhengzhou University (Natural Science), 2014, 0(3): 54-58 |
| |
| Authors: | MU Xiao-min ZHANG Zhong QU Yan LIU Zhan-wei |
| |
| Affiliation: | ( School of Information Engineering, Zhengzhou University, Zhengzhou 450001, China) |
| |
| Abstract: | There was a new image multiscale geometric analysis method based on the shearlet transformation. A facial expression recognition approach based on the shearlet transformation was explored, which was taken advantages of both the directional sensitivity and anisotropy of the shearlet transformation. The approach achieved 98.58% and 95.83% recognition rate respectively which was shown by experiments on the JAFFE database and RML database. The effectiveness of the proposed approach was verified by the result. |
| |
| Keywords: | shearlet transformation expression recognition support vector machine |
| 本文献已被 维普 等数据库收录！ |
|
