Intelligent object extraction algorithm based on foreground/background classification

Authors:
Jhing-Fa Wang;Han-Jen Hsu;Jyun-Sian Li
Affiliations:
Department of Electrical Engineering, National Cheng Kung University, Tainan, Taiwan;Department of Electrical Engineering, National Cheng Kung University, Tainan, Taiwan;Department of Electrical Engineering, National Cheng Kung University, Tainan, Taiwan
Venue:
EUC'05 Proceedings of the 2005 international conference on Embedded and Ubiquitous Computing
Year:
2005

Citing 11
Cited 1

An analysis of selected computer interchange color spaces

ACM Transactions on Graphics (TOG)
Image snapping

SIGGRAPH '95 Proceedings of the 22nd annual conference on Computer graphics and interactive techniques
Digital Image Processing

Digital Image Processing
Computer Vision

Computer Vision
Adaptive Split-and-Merge Segmentation Based on Piecewise Least-Square Approximation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Fragment-based image completion

ACM SIGGRAPH 2003 Papers
Inference of Segmented Color and Texture Description by Tensor Voting

IEEE Transactions on Pattern Analysis and Machine Intelligence
Lazy snapping

ACM SIGGRAPH 2004 Papers
"GrabCut": interactive foreground extraction using iterated graph cuts

ACM SIGGRAPH 2004 Papers
Poisson matting

ACM SIGGRAPH 2004 Papers
Improved techniques for automatic image segmentation

IEEE Transactions on Circuits and Systems for Video Technology

Extracting salient visual attention regions by color contrast and wavelet transformation

ISCIT'09 Proceedings of the 9th international conference on Communications and information technologies

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we propose an intelligent object extraction algorithm based on foreground/background classification. The proposed algorithm can offer the users more friendly interface for object extraction from image without unnecessary steps. After the interactive steps from user (marking the foreground and background parts), the wanted object is extracted from the background automatically. The proposed algorithm processes the input image by watershed to produce the regions. Then, the regions are labeled after marking parts of regions. We also introduce an implementation of hierarchical queues to store the unlabeled regions. The classification of foreground and background will generate the final image with selected object. In our experimental results, the proposed algorithm provides the output image with high efficiency. The wanted object is generated after user marking the foreground and background parts less than one second. In addition, the application of this work also can be used in image synthesis or object removal in other fields of image processing.