A scale- and orientation-adaptive extension of Local Binary Patterns for texture classification

作者：

Highlights：

• A scale- and rotation-invariant feature representation based on LBP is presented.

• LBPs are computed adaptively based on the estimated scale of an image.

• An estimate of global orientation is used to align LBP at a high angular resolution.

• Features independent of the intrinsic texture scale increase the adaptability.

• Significant improvements are achieved in scenarios with scaling and rotation.

摘要

Highlights•A scale- and rotation-invariant feature representation based on LBP is presented.•LBPs are computed adaptively based on the estimated scale of an image.•An estimate of global orientation is used to align LBP at a high angular resolution.•Features independent of the intrinsic texture scale increase the adaptability.•Significant improvements are achieved in scenarios with scaling and rotation.

论文关键词：LBP,Texture,Classification,Scale,Adaptive,Rotation,Invariant,Scale-space

论文评审过程：Received 10 September 2014, Revised 15 January 2015, Accepted 25 February 2015, Available online 4 March 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2015.02.024