时间: 2016年7月4日（周三）上午10:00-11:00

地点: 延长校区行健楼1018室

讲座: 颜色指导的深度增强

演讲者: 悉尼科技大学 Qiang Wu 副教授

演讲者简介：Qiang Wu received the B.Eng. and M.Eng. degrees from the Harbin Institute of Technology, Harbin, China, in 1996 and 1998 respectively and the P.h.D. degree from the University of Technology Sydney, Australia in 2004. He is now an associate professor in University of Technology Sydney. He is a core member of Global Big Data Technologies Centre. Dr. Qiang Wu’s research interests include computer vision, image processing, pattern recognition, machine learning, and multimedia processing. The application fields where the research outcomes are applied span over video security surveillance, biometrics, video data analysis, human-computer interaction. His research outcomes have been published in many premier international conferences including ECCV, CVPR, ICIP, and ICPR and the major international journals such as IEEE TIP, IEEE TSMC-B, IEEE TCSVT, IEEE TIFS, PR, PRL, Signal Processing, Signal Processing Letter. Dr. Qiang Wu is also a principal investigator and/or a tech. lead in several industry research projects collaborating with Toshiba, Microsoft, Nokia, Huawei, and Westpac Bank. Dr. Qiang Wu also serves several journals and conferences as reviewer including TPAMI, PR, TIP, TCSVT, TSMC-B, CVIU, IVC, PRL, Neurocomputing, IJPRAI and EURASIP Journal on Image and Video Processing. Dr. Qiang Wu has been involved in many international conferences and played different roles such as MMM2015, ICME2012, AVSS2010, MMSP2008, ISM2014, DICTA2010.

讲座摘要：

基于彩色图的深度图增强的基本假设是匹配的彩色图和深度图的边界具有一致性，然而，上述假设并不总是成立的。主流的基于马尔科夫随机场的方法通过调整平滑项的系数解决这个问题，但由于缺乏显式的匹配彩色图和深度图边界一致性测量，该类改进不能自适应控制匹配彩色图在深度图增强中的作用，因此，纹理拷贝和深度图边界模糊不可避免。我们提出一种显式的彩色深度图像对边界一致性度量方法，基于此方法的测量值被嵌入到马尔科夫随机场以自适应调整平滑项的系数。与十多个数据集的实验对比证明了该方法的稳定性和可靠性。