Multilevel thresholding of color images using globally informed artificial bee colony algorithm
Abstract Multilevel image thresholding presents a computational challenge as the number of thresholds increases, requiring efficient optimization techniques. The artificial bee colony (ABC) algorithm is a widely used metaheuristic for addressing this problem. Despite the good performance of the ABC...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-05238-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849769166037319680 |
|---|---|
| author | Ivona Brajević Jelena Ignjatović |
| author_facet | Ivona Brajević Jelena Ignjatović |
| author_sort | Ivona Brajević |
| collection | DOAJ |
| description | Abstract Multilevel image thresholding presents a computational challenge as the number of thresholds increases, requiring efficient optimization techniques. The artificial bee colony (ABC) algorithm is a widely used metaheuristic for addressing this problem. Despite the good performance of the ABC algorithm, it struggles with an inadequate balance between discovering new solutions and refining existing ones. This paper presents the globally informed artificial bee colony (giABC), an enhanced ABC variant, proposed for multilevel color image thresholding. To overcome the limitations of the ABC algorithm, giABC introduces two novel mutation operators. In the employed phase, solutions are dynamically guided toward the mean of the current better solutions, ensuring a sustained balance between global exploration and local enhancement. In the onlooker phase, solutions are further refined by combining attraction to the global best solution with adaptation to promising solutions, significantly enhancing both convergence speed and solution quality. The proposed giABC, along with the ABC, its two variants and the chaotically-enhanced Rao algorithm, were tested on twelve color images from the Berkeley dataset using Otsu’s objective function. Experimental results show that giABC outperforms the other metaheuristics in accuracy, robustness, peak signal-to-noise ratio and structural similarity index, with Wilcoxon signed-rank tests confirming its statistical significance. |
| format | Article |
| id | doaj-art-03a1a2c06a2e45c99d64e244a3dbf559 |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-03a1a2c06a2e45c99d64e244a3dbf5592025-08-20T03:03:32ZengNature PortfolioScientific Reports2045-23222025-07-0115111510.1038/s41598-025-05238-zMultilevel thresholding of color images using globally informed artificial bee colony algorithmIvona Brajević0Jelena Ignjatović1Faculty of Applied Management, Economics and Finance, University Business Academy in Novi SadFaculty of Science and Mathematics, University of NišAbstract Multilevel image thresholding presents a computational challenge as the number of thresholds increases, requiring efficient optimization techniques. The artificial bee colony (ABC) algorithm is a widely used metaheuristic for addressing this problem. Despite the good performance of the ABC algorithm, it struggles with an inadequate balance between discovering new solutions and refining existing ones. This paper presents the globally informed artificial bee colony (giABC), an enhanced ABC variant, proposed for multilevel color image thresholding. To overcome the limitations of the ABC algorithm, giABC introduces two novel mutation operators. In the employed phase, solutions are dynamically guided toward the mean of the current better solutions, ensuring a sustained balance between global exploration and local enhancement. In the onlooker phase, solutions are further refined by combining attraction to the global best solution with adaptation to promising solutions, significantly enhancing both convergence speed and solution quality. The proposed giABC, along with the ABC, its two variants and the chaotically-enhanced Rao algorithm, were tested on twelve color images from the Berkeley dataset using Otsu’s objective function. Experimental results show that giABC outperforms the other metaheuristics in accuracy, robustness, peak signal-to-noise ratio and structural similarity index, with Wilcoxon signed-rank tests confirming its statistical significance.https://doi.org/10.1038/s41598-025-05238-zMultilevel color image thresholdingOtsu methodArtificial bee colony algorithmOptimization |
| spellingShingle | Ivona Brajević Jelena Ignjatović Multilevel thresholding of color images using globally informed artificial bee colony algorithm Scientific Reports Multilevel color image thresholding Otsu method Artificial bee colony algorithm Optimization |
| title | Multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| title_full | Multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| title_fullStr | Multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| title_full_unstemmed | Multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| title_short | Multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| title_sort | multilevel thresholding of color images using globally informed artificial bee colony algorithm |
| topic | Multilevel color image thresholding Otsu method Artificial bee colony algorithm Optimization |
| url | https://doi.org/10.1038/s41598-025-05238-z |
| work_keys_str_mv | AT ivonabrajevic multilevelthresholdingofcolorimagesusinggloballyinformedartificialbeecolonyalgorithm AT jelenaignjatovic multilevelthresholdingofcolorimagesusinggloballyinformedartificialbeecolonyalgorithm |