Hierarchical semantic model and scattering mechanism based PolSAR image classification

作者：

Highlights：

• A hierarchical semantic model, consisting of a polarimetric sketch map and a region map, is proposed to partition a PolSAR image into hybrid, linear and homogenous region types.

• Complex image partition problem is translated into a sketch line grouping problem by defining some graph rules.

• A new classification framework is proposed based on the hierarchical semantic model. Different classification schemes are adopted according to the image characteristics of the three region types.

摘要

Highlights•A hierarchical semantic model, consisting of a polarimetric sketch map and a region map, is proposed to partition a PolSAR image into hybrid, linear and homogenous region types.•Complex image partition problem is translated into a sketch line grouping problem by defining some graph rules.•A new classification framework is proposed based on the hierarchical semantic model. Different classification schemes are adopted according to the image characteristics of the three region types.

论文关键词：Hierarchical semantic model,Polarimetric sketch map,Region map,Semantic-polarimetric classification

论文评审过程：Received 30 August 2015, Revised 24 February 2016, Accepted 25 February 2016, Available online 4 March 2016, Version of Record 23 August 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.02.020