Title :
Fast detect shadow from single outdoor image
Author :
Zhiping Xu ; Jinhong Pan
Author_Institution :
Sch. of Comput. Sciecne, Fudan Univ., Shanghai, China
Abstract :
There have been important advances in shadow detection in the images and videos recently. However, most shadow detection methods designed for the image sequences are not suitable for the single still image. The state-of-art shadow detection for still image is facing problems of not effective in general application. This paper presents a fast shadow detection method for single out-door image. The method is divided into three phase - image partition phase, partition enhance phase and shadow detection phase. By utilizing common phenomena in the shadow aspect, shadow detection in the single outdoor image becomes more simpler and more effective based on our research, and it runs at least 4-5 times faster than the instinct image based shadow detection method according to our experiments.
Keywords :
image sequences; object detection; image partition phase; image sequences; partition enhance phase; shadow detection method; shadow detection phase; single still outdoor image; Feature extraction; Humans; Image color analysis; Image sequences; Lighting; Neurons; Videos;
Conference_Titel :
Natural Computation (ICNC), 2011 Seventh International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-9950-2
DOI :
10.1109/ICNC.2011.6022404