An improved particle swarm optimization for multilevel thresholding medical image segmentation DOI Creative Commons
Jiaqi Ma, Jianmin Hu

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(12), P. e0306283 - e0306283

Published: Dec. 31, 2024

Multilevel thresholding image segmentation is one of the widely used methods, and it also an important means medical preprocessing. Replacing original costly exhaustive search approach, swarm intelligent optimization algorithms are recently to determine optimal thresholds for image, images tend have higher bit depth. Aiming at drawbacks premature convergence existing high-bit depth segmentation, this paper presents a pyramid particle based on complementary inertia weights (CIWP-PSO), Kapur entropy employed as objective. Firstly, according fitness value, divided into three-layer structure. To accommodate larger range caused by depth, particles in layer with worst value random opposition learning strategy. Secondly, pair introduced balance capability exploitation exploration. In part experiments, nine benchmark test CIWP-PSO effectiveness. Then, group Brain Magnetic Resonance Imaging (MRI) 12-bit utilized validate advantages compared other algorithms. According experimental results, optimized could achieve entropy, multi-level algorithm outperforms similar segmentation. Besides, we quality metrics evaluate impact different images, results show that MRI segmented has achieved best more times than comparison terms Structured Similarity Index Feature Index, which explains quality.

Language: Английский

A Comprehensive Survey of Multi-Level Thresholding Segmentation Methods for Image Processing DOI

Mohammad Amiriebrahimabadi,

Zhina Rouhi,

N. Mansouri

et al.

Archives of Computational Methods in Engineering, Journal Year: 2024, Volume and Issue: 31(6), P. 3647 - 3697

Published: March 27, 2024

Language: Английский

Citations

16
Multi-strategy learning-based particle swarm optimization algorithm for COVID-19 threshold segmentation DOI
Donglin Zhu, Jiaying Shen,

Yangyang Zheng

et al.

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 176, P. 108498 - 108498

Published: April 30, 2024

Language: Английский

Citations

5
A multi-threshold image segmentation method based on arithmetic optimization algorithm: A real case with skin cancer dermoscopic images DOI Creative Commons
Shuhui Hao, Changcheng Huang, Yi Chen

et al.

Journal of Computational Design and Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 14, 2025

Abstract Multi-threshold image segmentation (MTIS) is a crucial technology in processing, characterized by simplicity and efficiency, the key lies selection of thresholds. However, method's time complexity will grow exponentially with number To solve this problem, an improved arithmetic optimization algorithm (ETAOA) proposed paper, optimizer for optimizing process merging appropriate Specifically, two strategies are introduced to optimize optimal threshold process: elite evolutionary strategy (EES) tracking (ETS). First, verify performance ETAOA, mechanism comparison experiments, scalability tests, experiments nine state-of-the-art peers executed based on benchmark functions CEC2014 CEC2022. After that, demonstrate feasibility ETAOA domain, were performed using ten advanced methods skin cancer dermatoscopy datasets under low high thresholds, respectively. The above experimental results show that performs outstanding compared functions. Moreover, domain has superior conditions.

Language: Английский

Citations

0
Chaos-enhanced metaheuristics: classification, comparison, and convergence analysis DOI Creative Commons
Abdelhadi Limane, Farouq Zitouni, Saad Harous

et al.

Complex & Intelligent Systems, Journal Year: 2025, Volume and Issue: 11(3)

Published: Feb. 19, 2025

Language: Английский

Citations

0
Deep learning assisted optimal dispatch for renewable-based energy system considering consumer incentive scheme DOI
Mantosh Kumar, Kumari Namrata, Akshit Samadhiya

et al.

Cluster Computing, Journal Year: 2025, Volume and Issue: 28(4)

Published: Feb. 25, 2025

Language: Английский

Citations

0
Symmetric cross-entropy multi-threshold color image segmentation based on improved pelican optimization algorithm DOI Creative Commons
Chuang Zhang,

Yue-Han Pei,

Xiaoxue Wang

et al.

PLoS ONE, Journal Year: 2023, Volume and Issue: 18(6), P. e0287573 - e0287573

Published: June 29, 2023

To address the problems of low accuracy and slow convergence traditional multilevel image segmentation methods, a symmetric cross-entropy thresholding method (MSIPOA) with multi-strategy improved pelican optimization algorithm is proposed for global tasks. First, Sine chaotic mapping used to improve quality distribution uniformity initial population. A spiral search mechanism incorporating sine cosine improves algorithm's diversity, local pioneering ability, accuracy. levy flight strategy further ability jump out minima. In this paper, 12 benchmark test functions 8 other newer swarm intelligence algorithms are compared in terms speed evaluate performance MSIPOA algorithm. By non-parametric statistical analysis, shows greater superiority over algorithms. The then experimented threshold segmentation, eight images from BSDS300 selected as set MSIPOA. According different metrics Fridman test, outperforms similar cross entropy can be effectively applied

Language: Английский

Citations

9
Ultra‐widefield and high‐speed spiral laser scanning OR‐PAM: System development and characterization DOI Creative Commons
Mohsin Zafar, Rayyan Manwar, Laura Stone McGuire

et al.

Journal of Biophotonics, Journal Year: 2023, Volume and Issue: 16(7)

Published: March 31, 2023

Abstract Photoacoustic microscopy (PAM) is a high‐resolution imaging modality that has been mainly implemented with small field of view applications. Here, we developed fast PAM system utilizes unique spiral laser scanning mechanism and wide acoustic detection unit. The can image an area 12.5 cm 2 in 6.4 s. characterized using highly detailed phantoms. Finally, the capabilities were further demonstrated by sheep brain ex vivo rat vivo.

Language: Английский

Citations

8
A cross entropy and whale optimization algorithm based image segmentation for aerial images DOI
Saifuddin Ahmed, Anupam Biswas

International Journal of Information Technology, Journal Year: 2024, Volume and Issue: unknown

Published: April 30, 2024

Language: Английский

Citations

2
Multilevel Threshold Segmentation of Skin Lesions in Color Images Using Coronavirus Optimization Algorithm DOI Creative Commons
Yousef S. Alsahafi,

Doaa S. Elshora,

Ehab R. Mohamed

et al.

Diagnostics, Journal Year: 2023, Volume and Issue: 13(18), P. 2958 - 2958

Published: Sept. 15, 2023

Skin Cancer (SC) is among the most hazardous due to its high mortality rate. Therefore, early detection of this disease would be very helpful in treatment process. Multilevel Thresholding (MLT) widely used for extracting regions interest from medical images. paper utilizes recent Coronavirus Disease Optimization Algorithm (COVIDOA) address MLT issue SC images utilizing hybridization Otsu, Kapur, and Tsallis as fitness functions. Various are utilized validate performance proposed algorithm. The algorithm compared following five meta-heuristic algorithms: Arithmetic (AOA), Sine Cosine (SCA), Reptile Search (RSA), Flower Pollination (FPA), Seagull (SOA), Artificial Gorilla Troops Optimizer (GTO) prove superiority. all algorithms evaluated using a variety measures, such Mean Square Error (MSE), Peak Signal-To-Noise Ratio (PSNR), Feature Similarity Index Metric (FSIM), Normalized Correlation Coefficient (NCC). results experiments that surpasses several competing terms MSE, PSNR, FSIM, NCC segmentation metrics successfully solves issue.

Language: Английский

Citations

4
Fully Automatic Grayscale Image Segmentation: Dynamic Thresholding for Background Adaptation, Improved Image Center Point Selection, and Noise-Resilient Start/End Point Determination DOI Creative Commons
Junyan Li,

Xuewen Gui

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(20), P. 9303 - 9303

Published: Oct. 12, 2024

As the requirement for image uploads in various systems continues to grow, segmentation has become a critical task subsequent operations. Balancing efficiency and accuracy of is persistent challenge. This paper focuses on threshold-based grayscale methods proposes fully automated approach. The approach begins with implementation an improved OTSU algorithm determine optimal dynamic threshold, enabling process adjust adaptively varying backgrounds. A novel method selecting center points introduced address issue poor when point falls outside foreground area. To further enhance algorithm’s generalization capability accuracy, continuity detection-based developed start end foreground. Compared traditional algorithms, tests sample images four different scales revealed that proposed achieved average improvements precision, recall rates 14.97%, 1.28%, 17.33%, respectively, processing speed remaining largely unaffected. Ablation experiments validated effectiveness using strategy combinations, combination all three strategies resulting significant by 15.51% 16.72%, respectively.

Language: Английский

Citations

1