基于小波包变换的模糊判决纹理分类 |
| |
引用本文: | 王晓丹,吴崇明,靳华,赵荣椿.基于小波包变换的模糊判决纹理分类[J].数据采集与处理,2001,16(1):37-41. |
| |
作者姓名: | 王晓丹 吴崇明 靳华 赵荣椿 |
| |
作者单位: | 1. 西北工业大学计算机科学与工程系 2. 空军工程大学导弹学院计算机工程系 |
| |
基金项目: | 国家自然科学基金及国家教委开放实验室基金资助项目。 |
| |
摘 要: | 提出一种基于小波包变换的模糊判决纹理分类方法,采用图象的完全树结构小波变换提取多分辨率纹理特征,模糊判决分类器通过引入隶属度函数对待征模糊化,反映了各类纹理样本间存在的差异及随机噪声等畸变因素赞成的抽取特征值存在的不确定性,提高了纹理分类算法对噪声或畸变的鲁棒性,通过实验获得了较满意的结果。
|
关 键 词: | 纹理分类 小波包变换 模糊判决 图像处理 计算机 |
文章编号: | 1004-9037(2001)01-0037-05 |
修稿时间: | 2000年2月25日 |
Texture Classification by
CompletedWavelet Packet Transform and Fuzzy Decision |
| |
Abstract: | A new texture classification approach by completed wavelet packet transform and fuzzy decision is proposed. The multi-resolution character of the texture is extracted by completed wavelet packet transform. By introducing membership function of the feature, the fuzzy decision classifier can make decision by the membership. The introduced membership function reflects the uncertainty of the feature values made by the difference between samples made by random noise. The proposed approach increases the robustness to noise. Simulational experiment gives a satisfactory result. |
| |
Keywords: | texture classification tree-structured wavelet transform fuzzy decision |
本文献已被 CNKI 维普 万方数据 等数据库收录! |