Abstract | ||
---|---|---|
In this paper, we introduce a novel shadow removal technique that produces a shadow-free scene. There have been few studies concerning shadow removal, and the existing approaches cannot perfectly restore the original background patterns after removing shadows. With an acceptable number of differently illuminated images, the proposed algorithm simulates an artificial infinite il- luminant plane over the field of view. By employing the offset reduction technique, the constancy of the brightness is also reliably guaranteed. Fi- nally, a shadowless image without loss of textural details is obtained without any region extraction phase. Experimental results show that the method could successfully remove all of the visible shadows. The benefits of the proposed algorithm compared to the conventional shadow detection algorithms are the lower computational costs and the improved reliability. |
Year | Venue | Keywords |
---|---|---|
2002 | BMVC | field of view |
Field | DocType | Citations |
Field of view,Computer vision,Shadow,Pattern recognition,Computer science,Artificial intelligence,Standard illuminant,Active illumination,Offset (computer science),Brightness | Conference | 7 |
PageRank | References | Authors |
1.16 | 8 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
J. J. Yoon | 1 | 7 | 1.50 |
C. Koch | 2 | 10 | 2.21 |
Tim J. Ellis | 3 | 76 | 7.33 |