【海天学者学术报告】— Reinhard Klette “Concise Computer Vision”

学术报告I

报告题目：Introduction: Sharpness, Contrast (Saliency), Segmentation

报告人：Reinhard Klette 教授 （Auckland University of Technology）

报告时间：2015 年 10 月 30日（星期五）上午 10:00-12:00

报告地点：研教楼 208

校内联系人：苏志勋 教授       联系电话：84708351-8020

摘要: Introduce briefly into the notation used in the book “Concise Computer Vision”, Springer 2014, and report then about extensions towards the second edition of this book on the listed subjects. The lectures also address current research at the newly established Centre for Robotics

& Vision (CeRV) at AUT, Auckland.

学术报告II

报告题目：Multi-camera Systems and Stereo Matching

报告人：Reinhard Klette 教授 （Auckland University of Technology）

报告时间：2015 年 11 月 3日（星期二）下午 15:30-17:30

报告地点：研教楼 212

校内联系人：苏志勋 教授       联系电话：84708351-8020

摘要: These two lectures (1h each) address a standard technology (stereo analysis) in computer vision, with a particular focus on semi-global matching, a technology which was successfully applied in the group of Prof. Klette for vision-based driver assistence.

学术报告III

报告题目：Feature Detection and Tracking

报告人：Reinhard Klette 教授 （Auckland University of Technology）

报告时间：2015 年 11 月 6日（星期五）上午 10:00-12:00

报告地点：研教楼 208

校内联系人：苏志勋 教授       联系电话：84708351-8020

摘要: The two lectures (1 h each) start with low-level feature detction paradigms, recall the Lucas-Kanade tracker, and move then on to a discussion of particle filters and Kalman filters, illustrated for particular applications in vision-based driver assistance.

学术报告IV

报告题目：Object Detection

报告人：Reinhard Klette 教授 （Auckland University of Technology）

报告时间：2015 年 11 月 10日（星期二）下午 15:30-17:30

报告地点：研教楼 212

校内联系人：苏志勋 教授       联系电话：84708351-8020

摘要: These two lectures (1h each) introduce at first general measures for object detection. The Viola-Jones approach for appearance-based object detection (Haar-like features, AdaBoost) is discussed in detail, illustrated by vehicle driver monitoring. A second class of objects (pedestrians) is discussed for illustrating the use of a different object detection framework (random or Hough forests).

报告人介绍：

    Reinhard Klette is a professor at the School of Engineering at Auckland University of Technology.

    He has over 450 publications, among them close to 100 peer-reviewed journal papers, three text books (on image processing, computer vision for 3D shape recovery, and computer vision in general, just published in 2014), four monographs (on parallel computing, digital geometry, panoramic imaging, and shortest Euclidean paths), 24 edited research books and conference proceedings,  31 peer-reviewed book chapters, and about 300 peer-reviewed conference papers in the areas of computer vision, pattern recognition, computational photography, robotics, artificial intelligence, computational geometry, parallel computing, algorithm design, digital geometry and topology, and inductive inference of recursive functions.

His research interests are in areas of computer vision, ranging from basic theoretical fundamentals in digital geometry and topology to advanced subjects related to the 3-dimensional modelling and understanding of the real world based on recorded multi-ocular image and video data.

此次学术活动由学校国际化基金项目资助。