Improved bat algorithm applied to multilevel image thresholding
1Faculty of Mathematics, University of Sarajevo, 71000 Sarajevo, Bosnia And Herzegovina.
Thescientificworldjournal
|August 29, 2014
Summary
This study introduces an improved bat algorithm for multilevel image thresholding, enhancing image segmentation efficiency. The modified algorithm offers superior performance and faster convergence compared to existing methods.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Image Processing
Background:
- Multilevel image thresholding is crucial for image segmentation and higher-level processing.
- Exhaustive search for thresholds leads to exponential computational time increases.
- Swarm intelligence metaheuristics offer efficient solutions for complex optimization problems.
Purpose of the Study:
- To adapt the bat algorithm for multilevel image thresholding.
- To improve the bat algorithm's efficiency and effectiveness for image segmentation.
- To evaluate the performance of the enhanced bat algorithm against state-of-the-art methods.
Main Methods:
- Adjusting the standard bat algorithm for multilevel image thresholding.
- Integrating elements from differential evolution and artificial bee colony algorithms.
- Testing the improved bat algorithm on standard benchmark images.
Main Results:
- The standard bat algorithm shows comparable performance to other state-of-the-art algorithms.
- The improved bat algorithm outperforms five other state-of-the-art algorithms.
- The proposed algorithm enhances result quality and significantly improves convergence speed.
Conclusions:
- The enhanced bat algorithm is a highly effective method for multilevel image thresholding.
- The integration of differential evolution and artificial bee colony concepts improves bat algorithm performance.
- This approach offers a promising solution for efficient and accurate image segmentation.
