Sökning: id:"swepub:oai:DiVA.org:liu-203118" >
Particle Swarm Opti...
Particle Swarm Optimizer Variants for Multi-level Thresholding: Theory, Performance Enhancement and Evaluation
-
- Das, Arunita (författare)
- Midnapore Coll Autonomous, India
-
- Sasmal, Buddhadev (författare)
- Midnapore Coll Autonomous, India; Midnapore City Coll, India
-
- Dhal, Krishna Gopal (författare)
- Midnapore Coll Autonomous, India
-
visa fler...
-
- Hussien, Abdelazim (författare)
- Linköpings universitet,Programvara och system,Tekniska fakulteten,Fayoum Univ, Egypt; Middle East Univ, Jordan; Appl Sci Private Univ, Jordan
-
- Naskar, Prabir Kumar (författare)
- Govt Coll Engn & Text Technol, India
-
visa färre...
-
(creator_code:org_t)
- 2024
- 2024
- Engelska.
-
Ingår i: Archives of Computational Methods in Engineering. - : SPRINGER. - 1134-3060 .- 1886-1784.
- Relaterad länk:
-
https://urn.kb.se/re...
-
visa fler...
-
https://doi.org/10.1...
-
visa färre...
Abstract
Ämnesord
Stäng
- Multilevel thresholding (MLT) has a significant impact in the realm of image segmentation. It is a simple and efficient method for image segmentation. However, as the number of thresholds increases, the computational complexity of MLT increases exponentially, and proper thresholding crucially depends on the utilized objective function. Swarm intelligence (SI) with the optimal objective function is used to enhance the efficacy of MLT image segmentation. Among the various SI techniques, Particle Swarm Optimization (PSO) and its variants are utilized extensively in literature to improve the performance of MLT. As a result, the objective of this study is to provide an updated survey on this topic. This study provides a comprehensive report of the classical PSO, and its improved variants, and their applications in MLT domains. However, like other SI techniques, PSO and its widely utilized variant called Darwinian PSO (DPSO) have the drawbacks of imbalanced exploration and exploitation and premature convergence. Therefore, along with the updated survey report, this study develops an efficient variant of the DPSO called Cauchy DPSO with Opposition Learning (CaDPSOOL). The Cauchy distribution and opposition learning (OL) have been incorporated to enhance the optimization capability by preventing premature convergence and maintaining the good balance between exploration and exploitation. Furthermore, a new hybrid objective function has also been designed by considering Otsu and Tsallis entropy in MLT domain for better segmentation results. The proposed CaDPSOOL with hybrid objective function has been employed for standard color image and hematopathology image segmentation domains. The results of this study show that the proposed model produces better results than other state-of-the-art MLT models in terms of segmentation quality metrics.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datorteknik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Engineering (hsv//eng)
Publikations- och innehållstyp
- ref (ämneskategori)
- art (ämneskategori)
Hitta via bibliotek
Till lärosätets databas