A Survey on Shadow Removal

PhD Qualifying Examination


Title: "A Survey on Shadow Removal"

by

Mr. Yao XIAO


Abstract:

Shadows are one of the most noticeable illumination effects in natural 
scenes. Although they provide strong perceptual cues of spatial and 
geometrical scene information, they are still problematic to many vision 
algorithms such as segmentation, tracking and object recognition. A shadow 
removal pre-process will result in improved performance of these tasks.

Most shadow removal algorithms share the same two stages: shadow detection 
and reconstruction. With a single image as input, shadow regions or shadow 
boundaries are firstly determined and located. Based on the detection 
results, the reconstruction is performed to recover a shadow-free image. 
According to whether an algorithm demands user inputs, we categorize 
existing shadow removal methods into fully-automatic and user-assisted.  A 
survey of representative algorithms, advantages and drawbacks of each 
category is presented during this article. In addition, a novel shadow 
removal method using depth cue is introduced and demonstrated.


Date:			Monday, 27 January 2014

Time:                   3:00pm - 5:00pm

Venue:                  Room 3501
                         lifts 25/26

Committee Members:	Prof. Chi-Keung Tang (Supervisor)
 			Prof. Chiew-Lan Tai (Chairperson)
 			Dr. Huamin Qu
 			Prof. Long Quan


**** ALL are Welcome ****