一种基于压缩感知理论的纹理分类方法

针对传统纹理分类方法计算复杂的问题，本文基于bag-of-words模型提出了一种简单、新奇的纹理分类方法。在特征提取阶段，使用NSCT滤波器对局部图像块进行映射投影，然后通过观测矩阵提取其随机测量值特征；在纹理分类阶段，直接将随机特征嵌入到bag-of-words环境，并且直接在压缩域内进行学习和分类。利用纹理图像的稀疏性，本文提出的特征提取方法简单，并且在性能和复杂度上都优于传统特征提取方法。最后使用CUReT数据库进行数值试验，并与patch、patch-MRF、MR8、LBP四种最经典的方法进行比对，本文方法在分类精度以及实时性上有重要的改进。

A texture classification method based on the theory of compressed sensing

According to the theories of sparse representation and compressed sensing, this paper presents a simple, novel approach for texture classification based on bag-of-words model. At the feature extraction stage, a small set of random features is extracted from local image patches.The random features are embedded into a bag-of-words model to perform texture classification; thus, learning and classification are carried out in a compressed domain, yet by leveraging the sparse nature of texture images, our approach outperforms traditional feature extraction methods which involve careful design and complex steps. We have conducted extensive experiments on the CUReT databases, We show that our approach leads to significant improvements in classification accuracy and instantaneity.