Threshold image segmentation based on improved sparrow search algorithm

Wu, Dongmei; Yuan, Chengzhi

doi:10.1007/s11042-022-13073-x

Cited by 24 publications

(14 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, ten common test photos with varied histograms are subjected to HDAFAA comparison between the proposed methods for OTSU and Kapur's methods, as well as with recent benchmark optimization algorithms such as MTEMO, GA, PSO, and BF. Quantitative results show that HDAFA is highly efficient in terms of PSNR, mean, thresholds, iteration counts, and image segmentation quality [12]. Dongmei Wu and Chengzhi Yuan.…”

Section: Shikai Wang Et Al( 2021)mentioning

confidence: 99%

A Multilevel Image Thresholding Based on Hybrid Jaya Algorithm and Simulated Annealing

Abd,

Hussein Alrawi,

Bassel

2023

JUAPS

View full text Add to dashboard Cite

Section: Shikai Wang Et Al( 2021)mentioning

confidence: 99%

A Multilevel Image Thresholding Based on Hybrid Jaya Algorithm and Simulated Annealing

Abd,

Hussein Alrawi,

Bassel

2023

JUAPS

View full text Add to dashboard Cite

“…through the comparison of the above data, it can be found that each improved strategy has a certain degree of assistance to the improvement of the seeking performance of the algorithm, and the effectiveness of the different strategies is verified. The threshold selection criterion of the 2D maximum entropy (2D ME) algorithm is such that the total entropy of the segmented target and the background is maximized [22]. In the 2D segmentation method, not only each gray level information of the image is considered, but also the gray level information of its neighbors.…”

Section: Analysis Of the Effectiveness Of Improvement Strategiesmentioning

confidence: 99%

A mixed strategy improved dung beetle optimization algorithm and its application

Chang

Luo

Zhang

et al. 2023

Preprint

View full text Add to dashboard Cite

A mixed strategy improved dung beetle optimization (MSDBO) algorithm is proposed to address the problems of slow convergence speed, easy falling into local optimum, and insufficient search accuracy of the dung beetle optimization algorithm. Firstly, the good point set strategy is introduced to initialize the population and improve the population diversity. Then, the spiral search strategy is combined with the whale optimization algorithm to improve the location update of dung beetle reproduction and foraging behavior, balancing the local exploitation and global search ability of the algorithm, and improving the convergence ability of the algorithm. Finally, the Levy flight strategy is used to improve the location update of dung beetle stealing behavior and improve the algorithm's ability to jump out of the local the ability of the algorithm to jump out of local optimality. The results, tested with 12 benchmark functions and validated using the Wilcoxon rank sum test, show that the improved algorithm has significant advantages in terms of convergence speed, stability, and solution accuracy. In addition, we also applied the MSDBO algorithm to a two-dimensional maximum entropy image segmentation task, and the experimental results show that the MSDBO algorithm has good performance in image segmentation.

show abstract

“…algorithm 25 and Threshold segmentation algorithm 26 are compared experimentally using the abnormal mammogram images. Table 8 shows the comparative analysis of various segmentation methods in terms of Se, Sp, and CSA.…”

Section: Sensitivity Sementioning

confidence: 99%

Classifications of benign and malignant mammogram images using Gabor‐modified CNN architecture

Swetha

Vadivu

2023

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Breast cancer is one of the leading life killing cancers in women patients around the world. Digital mammogram is used to detect and segment the abnormal mass portions in breast region. In this article, the breast cancer regions are detected and segmented using the Gabor transform based Convolutional Neural Networks (CNN) classification approach. This proposed breast cancer detection method consist of the following modules Kirsch's edge detector, Gabor transform, CNN classification and Segmentation. The cancer pixels and healthy pixels are differentiated in edge pixels. Hence, the Kirsch's edge detector is used to detect the fine edge boundary pixelsthe in source mammogram image. Then, the spatial edge detected mammogram image is converted into multi resolution mammogram image using Gabor transform. This transformed multi resolution mammogram image is classified into Normal, Benign or Malignant in this article using the proposed CNN architecture. The developed CNN structure extracts more complex feature maps from their internal layers to improve the classification rate. The conventional and proposed CNN structure is differed with respect to the internal layers. The proposed CNN structure is optimized by reducing the usage of the internals and increases Convolutional Filters (CF) in each Convolutional Layer during utilization design process. After classification process over, the cancer regions in the abnormal mammogram images are segmented using morphological functions. The proposed breast cancer detection is evaluated on Mammographic Image Analysis Society (MIAS) dataset and their cancer region segmentation results are 98.4% of sensitivity, 98.9% of specificity and 99.2% of cancer region segmentation accuracy. The average Detection Rate (DR) of the proposed breast cancer detection method is about 98.3% on the set of mammogramthe images from MIAS dataset. These simulation results are compared with other state‐of‐the art methods and cross verified by the k‐fold verification algorithm.

show abstract

Threshold image segmentation based on improved sparrow search algorithm

Cited by 24 publications

References 61 publications

A Multilevel Image Thresholding Based on Hybrid Jaya Algorithm and Simulated Annealing

A Multilevel Image Thresholding Based on Hybrid Jaya Algorithm and Simulated Annealing

A mixed strategy improved dung beetle optimization algorithm and its application

Classifications of benign and malignant mammogram images using Gabor‐modified CNN architecture

Contact Info

Product

Resources

About