报告题目：Display of Polarization Data for Remote Sensing Tasks

报告时间：2018年6月20日(周三)下午14:30

报告地点：仲英楼（理学院）B-253

报告人：J. SCOTT TYO

报告摘要:Polarization data has emerged as a complementary strategy to spectral imaging for many remote sensing tasks. While multispectral and hyperspectral images provide details about the material makeup of a scene, polarization tends to provide data on the shape, orientation, and surface quality of objects. In this talk, we will discuss the basics of imaging polarimetry and many applications for which polarization sensing is important. The talk will continue to present recent research on the display of polarization data. We discuss new strategies for mapping polarization data into color that are designed to accomplish particular tasks such as classification, target detection, and shape recognition.

报告人简介：J. Scott Tyo教授，国际著名偏振光学专家，澳大利亚新南威尔士大学工程与信息技术学院院长，国际电气和电子工程师协会（IEEE）会士，美国光学学会（OSA）会士，国际光学工程学会（SPIE）会士。曾任美国亚利桑那大学（University of Arizona）光学与电子计算机工程教授，2014年被授予SPIE GG Stokes奖。在国际著名学术期刊Optics Letters、Optics Express、Applied Optics、IEEE Trans. AntennasPropagate、IEEE Trans. Geosci.、Remote Sens等发表学术研究论文200余篇，被国际同行他引2000余次，单篇最高引用600余次。

报告题目：Multisensor data integration for maximizing information retrieval

报告时间：2018年6月20日(周三)下午16:30

报告地点：仲英楼（理学院）B-253

报告人：Xiuping Jia

报告摘要:With the recent advances in remote sensing technologies for Earth Observation, satellite images from multiple sensors are widely available. Examples include multispectral/hyperspectral images, LiDAR and SAR data, as well as night-time imagery. They provide complementary information for a better understanding of Earth’s features. Night-time imagery can especially reflect human activities, and together with daytime data, can provide improved monitoring of urbanization. However, information extraction from such big spatial data is a challenging job. Traditional machine learning methods may have limited performance due to the increased data dimensionality and the multimodal characteristics.

In this talk, multisource data registration, feature selection and feature extraction techniques will be presented and discussed. One class oriented classification scheme will be introduced and discussed.

报告人简介：Xiuping Jia博士，国际著名遥感数据分析、高光谱图像分类专家，澳大利亚新南威尔士大学工程与信息技术学院博导，哈尔滨工程大学、西安电子科技大学、中国石油大学的客座教授，中国农业信息技术工程技术研究中心兼职研究员，IEEE GRSS Symposium Award Committee主席，《遥感数字图像分析》（Springer-Verlag, 3rd(1999)）第四版著者，国际著名期刊IEEE Transactions on Geoscience and Remote Sensing副主编，Journal of Soils and Sediments编辑，2018年获多萝西绿色奖。主要从事遥感图像处理和空间数据分析领域的研究，发表学术论文190余篇。