| Beyza Akyıldız, Caner Özcan, İsmail Rakıp Karaş Graph Cuts in Image Segmentation: A Review |
|---|
| Abstract. Image segmentation is critical for computer vision as it is used in object detection, image analysis and other applications. The ability of graph cut algorithms to model image data and segmentation as an energy minimization problem, their ability to achieve globally optimal solutions and their versatility have made them powerful tools for segmentation. This review examines the theoretical foundations, practical applications and recent advances in the field of graph cut algorithms for image segmentation. Commonly used energy functions and the algorithms that drive the optimization process are analyzed. Recent developments in the field are reviewed, including interactive approaches, multi-label problems, and deep learning integration. The study reveals some of the challenges related to designing cost functions, optimizing algorithms and processing large-scale data. It also suggests some research directions that can be explored, such as deep learning, spatio-temporal information fusion and user feedback. Finally, this comprehensive review provides valuable insights for both practitioners and researchers and highlights the future potential of graph cuts to further advance image understanding in computer vision. |
| Keywords: Graph cut algorithms, Image segmentation, Energy minimization, Optimization, Future directions |
 Download PDF |
| DOI: https://doi.org/10.54381/itta2024.09 |