원문정보
초록
영어
Fuzzy Cube Granule Structure (FCGS) for image segmentation is proposed in the paper. Firstly, the atomic cube granule is represented as the vector including the YCbCr values of pixel of color image and radii 0. Secondly, the join operation between two cube granules is designed to obtain the larger cube granule. Thirdly, the FCGS is formed by the fuzzy inclusion measure defined by join operation and the user-defined granularity threshold . Global Consistency Error (GCE), Variation of Information (VI), Rand Index (RI) are used to evaluate the segmentations. Images selected from BSD300 are used to verify the feasibility of FCGS.
목차
Abstract
1. Introduction
2. Fuzzy Cube Granule Structures
2.1 Representation of Cube Granule
2.2 Operations Between Two Granules
2.3 Fuzzy Inclusion Measure
2.4 Fuzzy Cube Granule Structure
3. Evaluation of Segmentation
3.1. Global Consistency Error
3.2. Variation of Information
3.3. Rand Index
4. Experiments
5. Conclusion
Acknowledgments
References
1. Introduction
2. Fuzzy Cube Granule Structures
2.1 Representation of Cube Granule
2.2 Operations Between Two Granules
2.3 Fuzzy Inclusion Measure
2.4 Fuzzy Cube Granule Structure
3. Evaluation of Segmentation
3.1. Global Consistency Error
3.2. Variation of Information
3.3. Rand Index
4. Experiments
5. Conclusion
Acknowledgments
References