题 目(TITLE):Document Image Enhancement
讲 座 人(SPEAKER): Prof. Chew Lim Tan;National University of Singapore
主 持 人 (CHAIR): Prof. Chenglin Liu
时 间 (TIME): 15:00PM, May25 (Wednesday)
地 点 (VENUE): The Second Meeting Room, 13th Floor
报告摘要(ABSTRACT):
In this talk, I will present two groups of document image enhancement work in our Lab. The first group concerns document image binarization for which we have proposed two methods, one in using background estimation and the other in using local contrast. The two methods enabled us to win the DIBCO 2009 and HDIBCO 2010 contests. Furthermore, we have also proposed a self learning framework that allows us to improve the performance of reported document image binarization methods on badly degraded images. The second group deals with blurred images that may be caused by either motion or defocus. We have proposed an automatic image blurred region detection and classification technique. This is done by examining singular value information for each image pixel. The blur types (i.e. motion or defocus) are then determined based on certain alpha channel constraint that requires neither image deblurring nor blur kernel estimation.
报告人简介(BIOGRAPHY):
Chew Lim Tan is a Professor in the Department of Computer Science, School of Computing, National University of Singapore. He received his B.Sc. (Hons) degree in physics in 1971 from University of Singapore, his M.Sc. degree in radiation studies in 1973 from University of Surrey, UK, and his Ph.D. degree in computer science in 1986 from University of Virginia, U.S.A. His research interests include document image analysis and natural language processing. He has published more than 360 research publications in these areas. He is Associate Editor of ACM Transactions on Asian Language Information Processing, Pattern Recognition, and Editorial Board Member of International Journal on Document Analysis and Recognition. He is a member of Governing Board of the International Association of Pattern Recognition (IAPR). He is also a senior member of IEEE.
承办单位:模式识别国家重点实验室 |