Exploring illumination robust descriptors for human epithelial type 2 cell classification

作者：

Highlights：

• Two introduced features achieve superior performance on the HEp-2 cell classification task.

• We propose a novel Spatial Shape Index Descriptor (SSID) to capture spatial layout of the second-order texture structures.

• We utilize a multi-scale Local Orientation Adaptive Descriptor (LOAD) to the HEp-2 cell classification task.

• We introduce a new large-scale HEp-2 data set that contains 63,445 cell images from the I3A Task-2 data set.

摘要

Highlights•Two introduced features achieve superior performance on the HEp-2 cell classification task.•We propose a novel Spatial Shape Index Descriptor (SSID) to capture spatial layout of the second-order texture structures.•We utilize a multi-scale Local Orientation Adaptive Descriptor (LOAD) to the HEp-2 cell classification task.•We introduce a new large-scale HEp-2 data set that contains 63,445 cell images from the I3A Task-2 data set.

论文关键词：HEp-2 cell classification,Spatial shape index descriptor,Local orientation adaptive descriptor,Illumination robust descriptors

论文评审过程：Received 28 October 2015, Revised 21 May 2016, Accepted 28 May 2016, Available online 2 June 2016, Version of Record 16 June 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.05.032