Sparse representation with multi-manifold analysis for texture classification from few training images

作者：

Highlights：

• The prime pyramid expands the training dataset with less redundancy.

• Sparse representation provides a compact model for textural images in each class.

• Multi-manifold analysis improves discriminative power while mitigates overfitting.

• Reasonably high classification accuracy is achieved with very few training images.

摘要

•The prime pyramid expands the training dataset with less redundancy.•Sparse representation provides a compact model for textural images in each class.•Multi-manifold analysis improves discriminative power while mitigates overfitting.•Reasonably high classification accuracy is achieved with very few training images.

论文关键词：Texture classification,Sparse representation,Manifold learning,Multi-manifold analysis,Few training image

论文评审过程：Received 24 July 2013, Revised 17 May 2014, Accepted 5 July 2014, Available online 11 July 2014.

论文官网地址：https://doi.org/10.1016/j.imavis.2014.07.001