5月27日（周二）名师讲坛：环境信息学的超光谱成像
北京交通大学“与大师面对面”名师讲坛系列活动
（机电学院，2014年春季第14周）
【主 题】：Hyperspectral Imaging in Environmental Informatics环境信息学的超光谱成像
【主讲人】：周 峻研究员，澳大利亚格里菲斯大学
【时 间】：2014年5月27日（星期二）下午14：30-16：00
【地 点】：机械工程楼九层活动中心
【主 办】：研究生工作部
【承 办】：机电学院
【主讲人简介】：
Jun Zhou received the B.S. degree in Computer Science from Nanjing University of Science and Technology, China, in 1996, the M.S. degree in Computer Science from Concordia University, Canada, and the Ph.D. degree in computing science from the University of Alberta, Canada, in 2006. He joined the School of Information and Communication Technology in Griffith University as a tenured lecturer in June 2012. At the same time, he is an adjunct visiting fellow in the Research School of Computer Science at the Australian National University (ANU) and a visiting scientist in CSIRO. Prior to his appointment in Griffith University, he had been a research fellow in ANU, and a researcher at NICTA Canberra Lab. Dr Zhou was a winner of the Discovery Early Career Research Award from the Australian Research Council in 2012. His research interests are in pattern recognition, computer vision, hyperspectral imaging, and their applications to environmental informatics.
周峻于1996年在中国南京理工大学获得计算机科学学士学位，在加拿大康考迪亚大学获得计算机科学硕士学位，2006年在加拿大阿尔伯塔大学获得计算机科学博士学位。他在2012年6月成为了格里菲斯大学讲师。同时，他也是澳大利亚国立大学(ANU)计算机科学研究院和澳大利亚联邦科学与工业研究组织(CSIRO)的访问学者。在他任命于格里菲斯大学之前，他是ANU和澳大利亚国家信息通讯技术中心(NICTA)堪培拉实验室的研究员。2012年周峻荣获澳大利亚科研基金会优秀青年研究奖。他的研究兴趣在于模式识别、计算机视觉、超光谱成像，以及它们在环境信息学的应用。
【讲座简介】：
Comparing to grayscale, RGB and multi-spectral images that capture data in one, three, or several wavelength bands, hyperspectral imagery contains tens or hundreds of continuous bands that provide rich information on the spectral and spatial distribution of materials of the objects in a scene. This has opened great opportunity for environment and computer vision research which is heavily relied on the capacity and quality of images for object detection and image classification. In this talk, I will give an overview on the latest development of the hyperspectral imaging technology, and show the advantages of performing spectral-spatial image analysis for better scene understanding. Particular focus will be put on sparsity constrained hyperspectral image classification and unmixing, with their applications to land cover, plant, and soil analysis.
相较于灰度，RGB和在一个、三个、或者几个波段捕捉数据的多光谱图像，超光谱成像包含几十或者几百连续的波段，这能提供有关对象的材料在一个场景的光谱和空间分布丰富的信息。这为环境和计算机视觉研究提供了大好的机会，而这在很大程度上依赖于用于目标检测和图像分类的图像的容量和质量。在这次讲座中，我会给出超光谱成像技术最新发展的概述，并且展示为了更好地场景理解而执行的光谱-空间图像分析的优点。本次讲座的重点是稀疏约束的超光谱图像分类和分离，以及它们在土地覆盖、植物和土壤分析中的应用。