中国科学院自动化研究所   设为首页   加入收藏  联系我们
 
English
网站首页     实验室概况     研究队伍     组织机构     学术交流     科研成果     人才培养     开放课题     创新文化     资源共享     联系我们
    学术讲座

Colour in a New Light

模式识别系列讲座

Lecture Series in Pattern Recognition

 

    (TITLE)Colour in a New Light

(SPEAKER)Prof. Brian Funt (Simon Fraser University, Canada)

(CHAIR)Prof. Weiming Hu

     (TIME):May 8 (Thursday), 2014, 10:00 AM

    (VENUE):No.1 Conference Room (3rd floor), Intelligence Building

报告摘要(ABSTRACT):

The colour (i.e., RGB in the context of digital imaging) of an object depends very strongly on the light illuminating it. When the light is changed, the colours change. Work in the field of colour constancy tries to account for such changes; however, is it really possible to do so?  In this talk, I will discuss two difficulties related to current approaches to this problem. The first difficulty relates to estimating the colour of the light. The vast majority of illumination-estimation methods assume there is only one source of light in a scene, but this assumption is almost always violated. What are the consequences of this? The second difficulty concerns metamer mismatching. Metamer mismatching refers to the fact that since many different surface reflectances can be metameric under a given light (i.e., yield identical RGB) the colour of an object under a second light is only constrained to lie within a volume of possible RGB values. Results using a new algorithm for computing exact metamer mismatch volumes show that this volume is much larger than one might have guessed. What are implications of this for colour constancy? In the context of light and colour, I will also present a new hue descriptor that is more robust to changes in the lighting than the standard CIECAM02 hue.

报告人简介(BIOGRAPHY):

Dr. Brian Funt is a Professor of Computing Science at Simon Fraser University in Vancouver, Canada. He joined SFU in 1980 and has published extensively on colour in computer vision and colour in digital photography since then. He was a pioneer of colour in computer vision and an early recipient of the Marr Prize. Recently, he received the best paper award at the CGIV conference in Amsterdam ("Metamer Mismatch Volumes" with Logvinenko and Godau). He is on the board of the Canadian Colour Research Society and is the Canadian representative to the International Committee for Imaging Science.

友情链接
 
中科院自动化研究所 模式识别国家重点实验室 事业单位  京ICP备14019135号-3
NLPR, INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES