基于方向经验模式分解的纹理分类

摘要

提出一种纹理分类的新方法——基于方向经验模式分解的纹理分类.这个方法自适应地将图像分解为一个特定方向的IMF,然后分析IMF通过二维希尔伯特变换的方向频率和包络.对不同的自然纹理图像进行实验,并将结果与文献[1]的结果做比较.实验结果证明,本文方法的识别效果优于文献[1].%This paper presents a method for texture classification based on Directional Empirical Mode Decomposition (DEMD). The approach adaptively decomposes images into local narrow band ingredients-Intrinsic Mode Function (IMF) and extracts their features including directional frequency and envelopes. Experiments are conducted on different texture images . Compared to the literature [1], the effect of our method for texture recognition is obvious excelled it.