Research article Special Issues

Multilevel thresholding using a modified ant lion optimizer with opposition-based learning for color image segmentation

  • Multilevel thresholding has important research value in image segmentation and can effectively solve region analysis problems of complex images. In this paper, Otsu and Kapur's entropy are adopted among thresholding segmentation methods. They are used as the objective functions. When the number of threshold increases, the time complexity increases exponentially. In order to overcome this drawback, a modified ant lion optimizer algorithm based on opposition-based learning (MALO) is proposed to determine the optimum threshold values by the maximization of the objective functions. By introducing the opposition-based learning strategy, the search accuracy and convergence performance are increased. In addition to IEEE CEC 2017 benchmark functions validation, 11 state-of-the-art algorithms are selected for comparison. A series of experiments are conducted to evaluate the segmentation performance of the algorithm. The evaluation metrics include: fitness value, peak signal-to-noise ratio, structural similarity index, feature similarity index, and computational time. The experimental data are analyzed and discussed in details. The experimental results significantly demonstrate that the proposed method is superior over others, which can be considered as a powerful and efficient thresholding technique.

    Citation: Shikai Wang, Kangjian Sun, Wanying Zhang, Heming Jia. Multilevel thresholding using a modified ant lion optimizer with opposition-based learning for color image segmentation[J]. Mathematical Biosciences and Engineering, 2021, 18(4): 3092-3143. doi: 10.3934/mbe.2021155

    Related Papers:

    [1] Chuanda Cai, Changgen Peng, Jin Niu, Weijie Tan, Hanlin Tang . Low distortion reversible database watermarking based on hybrid intelligent algorithm. Mathematical Biosciences and Engineering, 2023, 20(12): 21315-21336. doi: 10.3934/mbe.2023943
    [2] Qichao Ying, Jingzhi Lin, Zhenxing Qian, Haisheng Xu, Xinpeng Zhang . Robust digital watermarking for color images in combined DFT and DT-CWT domains. Mathematical Biosciences and Engineering, 2019, 16(5): 4788-4801. doi: 10.3934/mbe.2019241
    [3] Shaozhang Xiao, Xingyuan Zuo, Zhengwei Zhang, Fenfen Li . Large-capacity reversible image watermarking based on improved DE. Mathematical Biosciences and Engineering, 2022, 19(2): 1108-1127. doi: 10.3934/mbe.2022051
    [4] Wenfa Qi, Wei Guo, Tong Zhang, Yuxin Liu, Zongming Guo, Xifeng Fang . Robust authentication for paper-based text documents based on text watermarking technology. Mathematical Biosciences and Engineering, 2019, 16(4): 2233-2249. doi: 10.3934/mbe.2019110
    [5] Shanqing Zhang, Xiaoyun Guo, Xianghua Xu, Li Li, Chin-Chen Chang . A video watermark algorithm based on tensor decomposition. Mathematical Biosciences and Engineering, 2019, 16(5): 3435-3449. doi: 10.3934/mbe.2019172
    [6] Kaimeng Chen, Chin-Chen Chang . High-capacity reversible data hiding in encrypted images based on two-phase histogram shifting. Mathematical Biosciences and Engineering, 2019, 16(5): 3947-3964. doi: 10.3934/mbe.2019195
    [7] Hongyan Xu . Digital media zero watermark copyright protection algorithm based on embedded intelligent edge computing detection. Mathematical Biosciences and Engineering, 2021, 18(5): 6771-6789. doi: 10.3934/mbe.2021336
    [8] Qiuling Wu, Dandan Huang, Jiangchun Wei, Wenhui Chen . Adaptive and blind audio watermarking algorithm based on dither modulation and butterfly optimization algorithm. Mathematical Biosciences and Engineering, 2023, 20(6): 11482-11501. doi: 10.3934/mbe.2023509
    [9] Haoyu Lu, Daofu Gong, Fenlin Liu, Hui Liu, Jinghua Qu . A batch copyright scheme for digital image based on deep neural network. Mathematical Biosciences and Engineering, 2019, 16(5): 6121-6133. doi: 10.3934/mbe.2019306
    [10] Rong Li, Xiangyang Li, Yan Xiong, An Jiang, David Lee . An IPVO-based reversible data hiding scheme using floating predictors. Mathematical Biosciences and Engineering, 2019, 16(5): 5324-5345. doi: 10.3934/mbe.2019266
  • Multilevel thresholding has important research value in image segmentation and can effectively solve region analysis problems of complex images. In this paper, Otsu and Kapur's entropy are adopted among thresholding segmentation methods. They are used as the objective functions. When the number of threshold increases, the time complexity increases exponentially. In order to overcome this drawback, a modified ant lion optimizer algorithm based on opposition-based learning (MALO) is proposed to determine the optimum threshold values by the maximization of the objective functions. By introducing the opposition-based learning strategy, the search accuracy and convergence performance are increased. In addition to IEEE CEC 2017 benchmark functions validation, 11 state-of-the-art algorithms are selected for comparison. A series of experiments are conducted to evaluate the segmentation performance of the algorithm. The evaluation metrics include: fitness value, peak signal-to-noise ratio, structural similarity index, feature similarity index, and computational time. The experimental data are analyzed and discussed in details. The experimental results significantly demonstrate that the proposed method is superior over others, which can be considered as a powerful and efficient thresholding technique.



    Pancreatic cancer is a malignancy with a very poor prognosis, as the number of its new cases almost equals to related deaths [1]. The major type of pancreatic tumors is pancreatic ductal adenocarcinoma (PDAC), where metastasis is present in most cases, leading to an extremely low 5-year survival rate [2]. PDAC samples can be classified into four major molecular subtypes including squamous, pancreatic progenitor, immunogenic, and aberrantly differentiated endocrine exocrine (ADEX) subtypes, and squamous subtype had the worse prognosis [3]. Meanwhile, it has been demonstrated that PDAC patient with isolated liver metastases exhibited worse overall survival when compared to those with other sites of metastases [4]. Thus, for patients with PDAC, it is of great importance to investigate molecular mechanism behind liver-metastatic behaviors and identify markers to assess the risks of metastasis.

    So far, several researches have focused on demonstrating the genomic changes behind metastatic PDAC. Unique gene expression profiles were uncovered between liver and peritoneal metastatic PDAC cell lines, suggesting that different mechanisms were involved when site-specific metastases were present [5]. Aberrant expression of Zeb1, EphA2, DPC4 and MKK4 was found to be associated with metastatic pancreatic cancers [6,7], and dysregulated PRDM14 expression was confirmed to affect liver metastasis in mice [8]. Notably, Zeb1 is a key transcription factor promoting epithelial-mesenchymal transition (EMT) in pancreatic cancer [9]. Moreover, metastatic PDAC is often accompanied with an immunosuppressive tumor microenvironment (TME), and evidences suggest that stemness and differentiation of pancreatic ductal epithelial cells were associated with inflammatory cytokines in the hepatic microenvironment [10]. For example, interleukin (IL)-1α, a cytokine that mostly secreted by macrophages, were found to induce metastasis in PDAC, while a recent study has reported an association between macrophage-derived granulin and CD8+ T cell infiltration in PDAC metastatic tumors [11]. Pharmacological depletion of macrophages was deemed effective in reducing metastatic behavior in mice with pancreatic cancer [12]. However, molecular mechanisms possibly underlying altered macrophage secretion and infiltration are seldom addressed, thus it is of great necessity to explore significantly dysregulated genes in varied types of immune cells in the metastatic TME of PDAC patients [13]. In the present study, we focused on exploring the RNA signatures of liver metastasis in PDAC, anticipated to identify a set of liver metastasis-related genes, evaluate their prognostic values, and reveal key cell types involved in liver metastasis of PDAC.

    We collected four gene expression datasets including The Cancer Genome Atlas [14] (TCGA), GSE151580, GSE71729 [15] and E-MTAB-6134 [16] from UCSC Xena (https://xena.ucsc.edu/), Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), and ArrayExpress (https://www.ebi.ac.uk/arrayexpress/), for the systematic data analysis, respectively. The detailed clinical characteristics of the PDAC samples were summarized in Table 1. The first two gene expression datasets were used to identify liver metastasis-related genes. The RNA-seq data were normalized to FPKM or TPM, and the microarray data were pre-normalized by previous studies. Moreover, we also collected the cell types and their specific marker genes from previous study [17].

    Table 1.  The clinical characteristics of the PDAC samples used for data analysis.
    Clinical variables TCGA (n = 182) GSE71729 (n = 145) E-MTAB-6134 (n = 309) GSE151580 (n = 13)
    TNM stage
    I 11.8% (21) 6.8% (21) 0% (0)
    II 84.3% (150) 93.2% (288) 0% (0)
    III 1.7% (3) 0% (0) 0% (0)
    IV 2.2% (4) 0% (0) 100% (13)
    Gender
    Male 54.7% (99) 57.9% (179)
    Female 45.3% (82) 42.1% (130)
    Overall survival
    Dead 51.9% (94) 67.2% (84) 62.8% (181)
    Alive 48.1% (87) 32.8% (41) 37.2% (107)
    Liver metastasis
    Yes 11.5% (21) 100% (13)
    No 88.5% (161) 0% (0)

     | Show Table
    DownLoad: CSV

    The two-sample comparison was conducted to identify differentially expressed genes, and the differential gene expression levels were test by t test and fold changes. The thresholds of P-value and log2 (fold change) were selected at 0.01 and 0.25.

    The Gene set overrepresentation enrichment analysis was used to test the degree of overlapping genes between two gene sets. The gene set enrichment analysis was employed to test whether the gene set of interest was expressed at the lowest or highest order of all the genes ranked by a specified statistic. The two analyses were implemented in R clusterProfiler [18] package.

    The liver metastasis score was calculated using single-sample gene set enrichment analysis (ssGSEA) [19], which defines an enrichment score that represents the degree of absolute enrichment of a gene set in each sample within a given data set. The enrichment score, a variant of Kolmogorov-Smirnov statistic, was proposed by the previous study [20]. The ssGSEA was implemented in R GSVA [21] package.

    The overall survival time and status, and the liver metastasis score were used as the response and predictor variables for the Cox proportional hazard regression model. The Cox model was fitted using R survival [22] package (https://cran.r-project.org/web/packages/survival/index.html), and visualized by R survminer package (https://cran.r-project.org/web/packages/survminer/index.html). The log-rank test was employed to test the difference of survival time between the two groups.

    The statistical analysis was implemented in R-4.0.2. The two-sample comparisons were performed by Wilcoxon rank-sum test or student-t test. Multiple-sample comparisons were performed by analysis of variance (ANOVA) or Kruskal-Wallis test. P-value of 0.05 was indicated as statistical significance.

    To identify the liver metastasis-related genes in PDAC, we conducted a systematic analysis of two gene expression datasets. Specifically, we compared the primary tissues of liver-metastatic PDAC patients with those of PDAC without liver metastasis using TCGA cohort [14] (n = 182). Moreover, the primary tissues of liver-metastatic PDAC patients were also compared with their adjacent normal tissues using GSE151580 cohort (n = 19). Totally, we identified 154 genes upregulated in primary tissues of liver-metastatic PDAC patients as compared with both primary tissues of PDAC patients without liver metastasis and adjacent normal tissues (P-value < 0.01 and log2 (fold change) > 0.25, Supplementary Table S1).

    As shown in Figure 1A and 1B, the 154 genes had significantly different expression profiles between those tissues, and were termed as liver metastasis-related genes. Furthermore, the gene set enrichment analysis was conducted to characterize the biological function of those liver metastasis-related genes, and epithelia mesenchymal transition (EMT) was significantly enriched by those genes (Figure 1C, Supplementary Table S2), indicating that EMT was significantly associated with liver metastasis in PDAC. In addition, estrogen response, p53, and glycolysis pathways were also enriched by those genes, further suggesting their potential roles in the liver metastasis of PDAC.

    Figure 1.  The genes associated with liver metastasis in pancreatic ductal adenocarcinoma (PDAC). The upregulated genes in primary PDAC tissues with liver metastasis, which was compared with adjacent normal tissues (A) and primary tissues of PDAC patients with liver metastasis (B). (C) The pathways enriched by the liver metastasis-associated genes in PDAC. The node size and color represent the number of genes within the pathway and adjusted P-value for enrichment. The gene ratio was calculated by dividing the total number of genes by the number of upregulated genes within the pathway.

    To further investigate whether the liver metastasis-related genes were regulated at epigenetic level, we conducted correlation analysis between the DNA methylation level within promoter regions and their corresponding target genes using TCGA cohort. Among the liver metastasis-related genes, 18 were predicted to be epigenetically regulated by their promoter DNA methylation at a stringent threshold (Figure 2A, correlation test, FDR < 0.05, n = 182). Notably, SFN (Stratifin), a gene encoding a cell cycle checkpoint protein [23], was reversely correlated with the DNA methylation levels of 8 sites, including cg06720467, cg07786675, cg11348165, cg13374701, cg13466284, cg14825555, cg17330303, and cg21950166, within the promoter (Figure 2B). Moreover, KRT19 (Keratin-19), a well-known marker for ductal cell [24], was also negatively correlated with its DNA methylation levels of 6 sites. These results suggested that DNA methylation acted as a pivotal role in the regulation of the liver metastasis-related genes.

    Figure 2.  The regulation of genes associated with liver metastasis by DNA methylation. (A) The liver metastasis-associated genes significantly regulated by DNA methylation. The color bars on the top represent the genes regulated by DNA methylation, and the width of the bars represent the number of CpG sites. The green and red colors in the heatmap represent the negative and positive correlation, respectively. The DNA methylation and gene expression of two well-recognized metastatic genes, SFN (B) and KRT19 (C). The x and y axis represent the beta value of methylation (probe-level) and log2-transformed FPKM (Fragment Per Kilo Million).

    To reveal the association of the liver metastasis-related genes with the clinical characteristics of PDAC, we derived a liver metastasis score (LMS) based on the liver metastasis-related genes for TCGA PDAC patients. Specifically, the LMS of each PDAC patient was estimated by single-sample gene set enrichment analysis (ssGSEA). The PDAC patients with liver metastasis had significantly higher LMS than those without liver metastasis (Figure 3A, t test, P-value < 0.0001). Moreover, the LMS was observed lower in patients with stage I as compared with those with stage II (Figure 3B, P-value < 0.05). Furthermore, the comparative analysis of the disease types and tumor grades revealed that the ductal and lobular type had a higher liver metastasis score than adenomas and adenocarcinoma (Figure 3C, P-value < 0.0001), and higher LMS was observed in PDAC patients with higher grade (Figure 3D, P-value < 0.05), suggesting that patients with ductal and lobular type or higher grade might have a higher risk of liver metastasis. In addition, there were no significant associations of LMS with tumor sites and gender (Figure 3E, F). These results demonstrated that the LMS derived from liver metastasis-related genes was closely associated with the clinical characteristics of disease type and tumor grade in PDAC.

    Figure 3.  The association of liver metastasis score (LMS) with clinical factors. The differential LMS levels of PDAC patients with different metastatic status, TNM stages, pathological classifications, grades, anatomical locations, and genders are displayed in (A–F).

    To evaluate the prognostic significance of the liver metastasis-related genes in PDAC, we built a Cox proportional hazard regression model based on the liver metastasis score (LMS). The samples in The Cancer Genome Atlas (TCGA) cohort were divided into high and low LMS groups by the LMS median. Specifically, the high LMS group showed significantly shorter overall survival than low LMS group (Figure 4A, log-rank test, P-value < 0.05). To further confirm the prognostic value of LMS, we also divided the samples from two public datasets, termed as GSE71729 [15] (n = 145) and E-MTAB-6134 [16] (n = 309), into high and low LMS groups, respectively. Consistently, the prognostic differences were also observed in the two datasets (Figure 4B, C, Table 2, P-value < 0.05). Collectively, these results indicated that the LMS might be a prognostic predictor for PDAC.

    Figure 4.  The association of liver metastasis score (LMS) with PDAC prognosis. The association of liver metastasis score (LMS) with PDAC prognosis is evaluated in three independent cohorts, TCGA (A), E-MTAB-6134 (B), and GSE71729 (C). The horizontal and vertical dashed lines represent the 50% survival rate and corresponding survival time. The number at risk means the number of samples survived before that time point. The number of censoring means the number of samples whose follow-up is censored at that time point.
    Table 2.  The hazard ratio and coefficient for LMS in the Cox model.
    Cohort log2 HR HR log2 HR standard error Z P-value
    TCGA −0.56 0.57 0.21 −2.66 7.92 × 10−3
    GSE71729 −0.60 0.55 0.22 −2.65 8.01 × 10−3
    E-MTAB-6134 −0.66 0.52 0.15 −4.37 1.27 × 10−5

     | Show Table
    DownLoad: CSV

    As the tumor metastasis was regulated by the cell types infiltrating into the tumor microenvironment and some specific tumor cells, we then investigated whether some specific cell types were responsible for liver metastasis in PDAC. Firstly, we collected the marker genes of cell types in PDAC tissues from earlier study [17]. The overrepresentation test of the liver metastasis-related genes revealed that those genes were highly enriched in the marker genes of malignant ductal cell (Figure 5A). Specifically, some metastasis-related genes such as MET (MET Proto-Oncogene), KRT19 (Keratin-19), KRT7 (Keratin-7), AGRN (Agrin), SDC4 (Syndecan-4), and SFN (Stratifin) were found to be specifically expressed in malignant ductal cell, indicating that the liver metastasis-related genes might be expressed in malignant ductal cells, which might be responsible for the liver metastasis in PDAC. Moreover, we also conducted correlation analysis between LMS and gene expression levels. Gene set enrichment analysis (GSEA) revealed that most of the marker genes in malignant ductal cell were positively correlated with LMS (Figure 5B, FDR < 0.05), further suggesting that malignant ductal cells were responsible for liver metastasis via expressing the liver metastasis-related genes. Furthermore, marker genes of M0 macrophage were also highly positively correlated with LMS (Figure 5C, FDR < 0.05). In addition, as motivated by previous study [25], we also estimated the relative abundance of macrophage M0 in PDAC tissues using gene expression and macrophage M0-specific genes by ssGSEA method. Consistently, the LMS was highly positively correlated with the relative activity of macrophage M0 (Figure 5D, Spearman correlation: 0.784). Therefore, the positive correlation between LMS and malignant ductal cell or macrophage M0 specific genes indicated that the two cell types might have the potential to promote liver metastasis of PDAC.

    Figure 5.  The association of metastasis-associated genes with malignant ductal cells and macrophage M0. (A) The shared number of genes between metastasis-associated genes and malignant ductal cell-specific genes. (B) Positive correlation of malignant ductal cell-specific genes with liver metastasis score (LMS). (C) Positive correlation of macrophage M0-specific genes with liver metastasis score (LMS). The genes were ranked by the correlation coefficients by decreasing order. The black lines in the middle represent the cell type specific genes. These cell-type specific genes were mainly clustered within the gene sets with high positive correlation with LMS (on the left). (D) The correlation between LMS and the relative abundance of macrophage M0.

    Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies worldwide [26]. Liver metastasis accounts for the high mortality rate and confers poor prognosis in PDAC [27]. Therefore, a better understanding of the mechanisms underlying the acquisition of the metastatic potential in PDAC is highly desirable.

    In this study, we conducted a systematic analysis of gene expression data to identify liver metastasis-related genes in PDAC. Specifically, a total of 154 genes were identified to be upregulated in primary tissues of PDAC with liver metastasis. These liver metastasis-related genes were involved in signaling pathways such as EMT, estrogen response, p53, and glycolysis. EMT, a developmental program that enables stationary epithelial cells to gain the ability to migrate and invade, has been implicated in carcinogenesis and cancer metastasis by enhancing mobility and invasion [28]. Consistently, EMT was also found as the key players in PDAC metastasis [29]. Particularly, glycolysis was also identified as a key pathway involved in liver metastasis of PDAC. The metabolic reprogramming from oxidative phosphorylation to glycolysis in pancreatic cancer cells, has been found to enhance the invasion-metastasis cascade by promoting EMT, angiogenesis and metastatic colonization of distant organs [30]. Combined with the result in this study, the liver metastasis-related genes were functionally relevant to liver metastasis in PDAC.

    The integrative analysis of DNA methylation and gene expression data revealed that the liver metastasis-related genes were primarily regulated at epigenetic level. Particularly, SFN, a cell cycle checkpoint protein, and KRT19, a marker gene for ductal cells, were predicted to be regulated by multiple methylation sites at the promoter. It should be noted that the two genes have been found as novel biomarkers of poor prognosis in PDAC [31,32]. We thus speculated that the overexpression of SFN might be associated with uncontrolled cell cycle progression and stemness maintenance, while the high expression of KRT19 gave us a hint that ductal cells might play key roles in liver metastasis in PDAC. In addition, some other DNA methylation-associated genes were also reported to be associated with tumor initiation or progression. For example, TACSTD2 (tumor associated calcium signal transducer 2) encoded a carcinoma-associated antigen [33], suggesting that it might be a potential diagnostic marker in PDAC.

    Clinically, we found that liver metastasis score (LMS), derived from liver metastasis-related genes, was found to be closely associated with clinical characteristics associated with prognostic outcomes [34], such as disease type and tumor grade, in PDAC. Furthermore, we also divided the samples from TCGA, GSE71729 and E-MTAB-6134 cohorts into high and low LMS groups, and the two groups exhibited significantly different prognostic outcomes across the three cohorts, suggesting that the LMS might be a promising biomarker for risk stratification in PDAC. However, as the liver metastasis-related genes used to derive LMS were mainly enriched in glycolysis and EMT pathway, which were altered in many other cancers, the LMS might not be the PDAC-specific prognostic marker.

    Furthermore, we also found that the liver metastasis-related genes were primarily expressed in malignant ductal cells by integrative analysis of the bulk and single-cell gene expression data. An earlier study [17] reported that the malignant ductal cells might be transited from ductal cells with abnormal gene expression profiles to malignant ductal cells by trajectory analysis. The malignant ductal cells were characterized to have highly proliferative and migratory subpopulations [17]. Moreover, the high correlation between LMS and M0 macrophage indicated that M0 macrophage might promote liver metastasis of PDAC. The macrophage cell subpopulations in tumors commonly refer to M2 macrophage, exhibiting anti-inflammatory and pro-tumoral effects [35], however, the role of unpolarized macrophage (M0) in tumor tissues was rarely reported. In accordance with our finding, low expression of M0 macrophage is associated with better clinical prognosis in bladder cancer patients [36]. Collectively, the malignant ductal cells and M0 macrophage might function as tumor-promoting cells in PDAC.

    However, this study still has some limitations. First, the biological function of some novel liver metastasis-related genes has not been validated by experiments. Second, the clinical significance of malignant ductal cells needs to be evaluated by clinical specimens. In addition, the protein expression levels of some important liver metastasis-related genes need to be evaluated. In summary, the systematic analysis greatly improves our understanding of the critical cell types and genes expressed by these cell types in the process of liver metastasis in PDAC.

    This project is supported by the Lead project of Western medicine of the Shanghai Science Committee (No. 16411967200), the PLA's youth training program of the medical science and technology (No. 16QNP095), and National Natural Science Foundation of China (No. 81871992).

    The authors declare that they have no conflicts of interest.



    [1] N. M. Zaitoun, M. J. Aqel, Survey on Image Segmentation Techniques, Procedia Comput. Sci., 65 (2015), 797-806. doi: 10.1016/j.procs.2015.09.027
    [2] M. Sridevi, C. Mala, A Survey on Monochrome Image Segmentation Methods, Procedia Technol., 6 (2012), 548-555. doi: 10.1016/j.protcy.2012.10.066
    [3] A. K. M. Khairuzzaman, S. Chaudhury, Multilevel thresholding using grey wolf optimizer for image segmentation, Expert Syst. Appl., 86 (2017), 64-76. doi: 10.1016/j.eswa.2017.04.029
    [4] J. Tang, Y. Wang, C. Huang, H. Liu, N. Al-Nabhan, Image edge detection based on singular value feature vector and gradient operator, Math. Biosci. Eng., 17 (2020), 3721-3735. doi: 10.3934/mbe.2020209
    [5] X. Song, Y. Wang, Q. Feng, Q. Wang, Improved graph cut model with features of superpixels and neighborhood patches for myocardium segmentation from ultrasound image, Infinite Study, 2019.
    [6] X. Lu, Z. You, M. Sun, J. Wu, Z. Zhang, Breast cancer mitotic cell detection using cascade convolutional neural network with U-Net, Math. Biosci. Eng., 18 (2021), 673-695. doi: 10.3934/mbe.2021036
    [7] H. Jia, K. Sun, W. Song, X. Peng, C. Lang, Y. Li, Multi-Strategy Emperor Penguin Optimizer for RGB Histogram-Based Color Satellite Image Segmentation Using Masi Entropy, IEEE Access, 7 (2019), 134448-134474. doi: 10.1109/ACCESS.2019.2942064
    [8] S. Wang, H. Jia, X. Peng, Modified salp swarm algorithm based multilevel thresholding for color image segmentation, Math. Biosci. Eng., 17 (2019), 700-724.
    [9] A. Dirami, K. Hammouche, M. Diaf, P. Siarry, Fast multilevel thresholding for image segmentation through a multiphase level set method, Signal Process., 93 (2013), 139-153. doi: 10.1016/j.sigpro.2012.07.010
    [10] E. Hamuda, M. Glavin, E. Jones, A survey of image processing techniques for plant extraction and segmentation in the field, Comput. Electron. Agric., 125 (2016), 184-199. doi: 10.1016/j.compag.2016.04.024
    [11] S. Kotte, R. K. Pullakura, S. K. Injeti, Optimal multilevel thresholding selection for brain MRI image segmentation based on adaptive wind driven optimization, Measurement, 130 (2018), 340-361. doi: 10.1016/j.measurement.2018.08.007
    [12] N. Otsu, A threshold selection method from gray-level histograms, IEEE Trans. Syst. Man Cybern., 9 (1979), 62-66. doi: 10.1109/TSMC.1979.4310076
    [13] J. N. Kapur, P. Sahoo, A. K. C. Wong, A new method for gray-level picture thresholding using the entropy of the histogram, Comput. Vis. Graph Image Process., 29 (1985), 273-285. doi: 10.1016/0734-189X(85)90125-2
    [14] A. K.Bhandari, V. K. Singh, A. Kumar, G. K. Singh, Cuckoo search algorithm and wind driven optimization based study of satellite image segmentation for multilevel thresholding using Kapur's entropy, Expert Syst. Appl., 41 (2014), 3538-3560. doi: 10.1016/j.eswa.2013.10.059
    [15] M. A. E. Aziz, A. A. Ewees, A. E. Hassanien, Whale Optimization Algorithm and Moth-Flame Optimization for multilevel thresholding image segmentation, Expert Syst. Appl., 83 (2017), 242-256. doi: 10.1016/j.eswa.2017.04.023
    [16] K. P. Baby Resma, M. S. Nair, Multilevel thresholding for image segmentation using Krill Herd Optimization algorithm, J. King Saud Univ. Comput. Inf. Sci., (2018), forthcoming.
    [17] A. Ibrahim, A. Ahmed, S. Hussein, A. E. Hassanien, Fish image segmentation using Salp Swarm Algorithm, in The International Conference on Advanced Machine Learning Technologies and Applications (AMLTA2018), Springer, (2018), 42-51.
    [18] S. Ouadfel, A. Taleb-Ahmed, Social spiders optimization and flower pollination algorithm for multilevel image thresholding: A performance study, Expert Syst. Appl., 55 (2016), 566-584. doi: 10.1016/j.eswa.2016.02.024
    [19] M. Díaz-Cortés, N. Ortega-Sánchez, S. Hinojosa, D. Oliva, E. Cuevas, R. Rojas, et al., A multi-level thresholding method for breast thermograms analysis using Dragonfly algorithm, Infrared Phys. Technol., 93 (2018), 346-361. doi: 10.1016/j.infrared.2018.08.007
    [20] S. C. Satapathy, N. Sri Madhava Raja, V. Rajinikanth, A. S. Ashour, N. Dey, Multi-level image thresholding using Otsu and chaotic bat algorithm, Neural Comput. Appl., 29 (2018), 1285-1307. doi: 10.1007/s00521-016-2645-5
    [21] M. Salvi, F. Molinari, Multi-tissue and multi-scale approach for nuclei segmentation in H & E stained images, BioMed. Eng. OnLine., 17 (2018), 89. doi: 10.1186/s12938-018-0518-0
    [22] Y. Feng, H. Zhao, X. Li, X. Zhang, H. Li, A multi-scale 3D Otsu thresholding algorithm for medical image segmentation, Digital Signal Process., 60 (2017), 186-199. doi: 10.1016/j.dsp.2016.08.003
    [23] D. Zhao, L. Liu, F. Yu, A. A. Heidari, M. Wang, G. Liang, et al., Chaotic random spare ant colony optimization for multi-threshold image segmentation of 2D Kapur entropy, Knowl. Based Syst., 216 (2021), 106510. doi: 10.1016/j.knosys.2020.106510
    [24] D. Zhao, L. Liu, F. Yu, A. A. Heidari, M. Wang, D. Oliva, et al., Ant colony optimization with horizontal and vertical crossover search: Fundamental visions for multi-threshold image segmentation, Expert Syst. Appl., 167 (2021), 114122. doi: 10.1016/j.eswa.2020.114122
    [25] L. He, S. Huang, An efficient krill herd algorithm for color image multilevel thresholding segmentation problem, Appl. Soft Comput., 89 (2020), 106063. doi: 10.1016/j.asoc.2020.106063
    [26] I. Hilali-Jaghdam, A. B. Ishak, S. Abdel-Khalek, A. Jamal, Quantum and classical genetic algorithms for multilevel segmentation of medical images: A comparative study, Comput. Commun., 162 (2020), 83-93. doi: 10.1016/j.comcom.2020.08.010
    [27] B. Wu, J. Zhou, X. Ji, Y. Yin, X. Shen, An ameliorated teaching-learning-based optimization algorithm based study of image segmentation for multilevel thresholding using Kapur's entropy and Otsu's between class variance, Inf. Sci., 533 (2020), 72-107. doi: 10.1016/j.ins.2020.05.033
    [28] S. Mirjalili, The Ant Lion Optimizer, Adv. Eng. Software, 83 (2015), 80-98. doi: 10.1016/j.advengsoft.2015.01.010
    [29] M. J. Hadidian-Moghaddam, S. Arabi-Nowdeh, M. Bigdeli, D. Azizian, A multi-objective optimal sizing and siting of distributed generation using ant lion optimization technique, Ain Shams Eng. J., 9 (2018), 2101-2109. doi: 10.1016/j.asej.2017.03.001
    [30] M. Raju, L. C. Saikia, N. Sinha, Automatic generation control of a multi-area system using ant lion optimizer algorithm based PID plus second order derivative controller, Int. J. Electr. Power Energy Syst., 80 (2016), 52-63. doi: 10.1016/j.ijepes.2016.01.037
    [31] P. Saxena, A. Kothari, Ant Lion Optimization algorithm to control side lobe level and null depths in linear antenna arrays, Int. J. Electron. Commun., 70 (2016), 1339-1349. doi: 10.1016/j.aeue.2016.07.008
    [32] E. Umamaheswari, S. Ganesan, M. Abirami, S. Subramanian, Cost Effective Integrated Maintenance Scheduling in Power Systems using Ant Lion Optimizer, Energy Procedia, 117 (2017), 501-508. doi: 10.1016/j.egypro.2017.05.176
    [33] P. D. P. Reddy, V. C. V. Reddy, T. G. Manohar, Ant Lion optimization algorithm for optimal sizing of renewable energy resources for loss reduction in distribution systems, J. Electr. Syst. Inf. Technol., 5 (2018), 663-680. doi: 10.1016/j.jesit.2017.06.001
    [34] D. Oliva, S. Hinojosa, M. A. Elaziz, N. Ortega-Sánchez, Context based image segmentation using antlion optimization and sine cosine algorithm, Multimedia Tools Appl., 77 (2018), 25761-25797. doi: 10.1007/s11042-018-5815-x
    [35] C. Jin, Z. Ye, L. Yan, Y. Cao, A. Zhang, L. Ma, et al., Image Segmentation Using Fuzzy C-means Optimized by Ant Lion Optimization, in 2019 10th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), IEEE, (2019), 388-393.
    [36] X. Yue, H. Zhang, A Novel Industrial Image Contrast Enhancement Technique Based on an Improved Ant Lion Optimizer, Arab J. Sci. Eng., 46 (2021), 3235-3246. doi: 10.1007/s13369-020-05148-4
    [37] Z. Wu, D. Yu, X. Kang, Parameter identification of photovoltaic cell model based on improved ant lion optimizer, Energy Convers. Manage., 151 (2017), 107-115. doi: 10.1016/j.enconman.2017.08.088
    [38] K. R. Subhashini, J. K. Satapathy, Development of an Enhanced Ant Lion Optimization Algorithm and its Application in Antenna Array Synthesis, Appl. Soft Comput., 59 (2017), 153-173. doi: 10.1016/j.asoc.2017.05.007
    [39] S. K. Majhi, S. Biswal, Optimal cluster analysis using hybrid K-Means and Ant Lion Optimizer, Karbala Int. J. Mod. Sci., 4 (2018), 347-360. doi: 10.1016/j.kijoms.2018.09.001
    [40] R. Sarkhel, N. Das, A. K. Saha, M. Nasipuri, An improved Harmony Search Algorithm embedded with a novel piecewise opposition based learning algorithm, Eng. Appl. Artif. Intell., 67 (2018), 317-330. doi: 10.1016/j.engappai.2017.09.020
    [41] A. A. Ewees, M. A. Elaziz, E. H. Houssein, Improved grasshopper optimization algorithm using opposition-based learning, Expert Syst. Appl., 112 (2018), 156-172. doi: 10.1016/j.eswa.2018.06.023
    [42] M. A. Ahandani, Opposition-based learning in the shuffled bidirectional differential evolution algorithm, Swarm Evol. Comput., 26 (2016), 64-85. doi: 10.1016/j.swevo.2015.08.002
    [43] N. H. Awad, M. Z. Ali, P. N. Suganthan, Ensemble sinusoidal differential covariance matrix adaptation with Euclidean neighborhood for solving CEC2017 benchmark problems, in 2017 IEEE Congress on Evolutionary Computation (CEC), IEEE, (2017), 372-379.
    [44] R. Roy, S. Laha, Optimization of Stego image retaining secret information using genetic algorithm with 8-connected PSNR, Procedia Comput. Sci., 60 (2015), 468-477. doi: 10.1016/j.procs.2015.08.168
    [45] A. Tanchenko, Visual-PSNR measure of image quality, J. Visual Commun. Image Represent., 25 (2014), 874-878. doi: 10.1016/j.jvcir.2014.01.008
    [46] Z. Wang, A. C. Bovik, H. R. Sheikh, E. P. Simoncelli, Image quality assessment: from error visibility to structural similarity, IEEE Trans. Image Process., 13 (2004), 600-612. doi: 10.1109/TIP.2003.819861
    [47] V. Bruni, D. Vitulano, An entropy based approach for SSIM speed up, Signal Process., 135 (2017), 198-209. doi: 10.1016/j.sigpro.2017.01.007
    [48] C. Li, A. C. Bovik, Content-partitioned structural similarity index for image quality assessment, Signal Process. Image Commun., 25 (2010), 517-526. doi: 10.1016/j.image.2010.03.004
    [49] L. Zhang, L. Zhang, X. Mou, D. Zhang, FSIM: A feature similarity index for image quality assessment, IEEE Trans. Image Process., 20 (2011), 2378-2386. doi: 10.1109/TIP.2011.2109730
    [50] J. John, M. S. Nair, P. R. A. Kumar, M. Wilscy, A novel approach for detection and delineation of cell nuclei using feature similarity index measure, Biocybern. Biomed. Eng., 36 (2016), 76-88. doi: 10.1016/j.bbe.2015.11.002
    [51] S. K. Dinkar, K. Deep, Opposition based Laplacian Ant Lion Optimizer, J. Comput. Sci., 23 (2017), 71-90. doi: 10.1016/j.jocs.2017.10.007
    [52] M. Wang, X. Zhao, A. A. Heidari, H. Chen, Evaluation of constraint in photovoltaic models by exploiting an enhanced ant lion optimizer, Sol. Energy, 211 (2020), 503-521. doi: 10.1016/j.solener.2020.09.080
    [53] The Berkeley Segmentation Dataset and Benchmark. Available from: https://www2.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/.
    [54] H. Jia, X. Peng, W. Song, C. Lang, Z. Xing, K. Sun, Multiverse Optimization Algorithm Based on Lévy Flight Improvement for Multithreshold Color Image Segmentation, IEEE Access, 7 (2019), 32805-32844. doi: 10.1109/ACCESS.2019.2903345
    [55] A. K. M. Khairuzzaman, S. Chaudhury, Masi entropy based multilevel thresholding for image segmentation, Multimed. Tools Appl., 78 (2019), 33573-33591. doi: 10.1007/s11042-019-08117-8
    [56] A. K. Bhandari, A. Kumar, G. K. Singh, Modified artificial bee colony based computationally efficient multilevel thresholding for satellite image segmentation using Kapur's, Otsu and Tsallis functions, Expert Syst. Appl., 42 (2015), 1573-1601. doi: 10.1016/j.eswa.2014.09.049
    [57] S. Kotte, P. R. Kumar, S. K. Injeti, An efficient approach for optimal multilevel thresholding selection for gray scale images based on improved differential search algorithm, Ain Shams Eng. J., 9 (2018), 1043-1067. doi: 10.1016/j.asej.2016.06.007
    [58] V. K. Bohat, K. V. Arya, A new heuristic for multilevel thresholding of images, Expert Syst. Appl., 117 (2019), 176-203. doi: 10.1016/j.eswa.2018.08.045
    [59] A. K. Bhandari, A novel beta differential evolution algorithm-based fast multilevel thresholding for color image segmentation, Neural Comput. Appl., 32 (2020), 4583-4613. doi: 10.1007/s00521-018-3771-z
    [60] D. H. Wolpert, W. G. Macready, No free lunch theorems for optimization, IEEE Trans. Evol. Comput., 1 (1997), 67-82. doi: 10.1109/4235.585893
  • This article has been cited by:

    1. Tina Draškovič, Nina Zidar, Nina Hauptman, Circulating Tumor DNA Methylation Biomarkers for Characterization and Determination of the Cancer Origin in Malignant Liver Tumors, 2023, 15, 2072-6694, 859, 10.3390/cancers15030859
    2. Tao Liu, Jian Chen, An-an Liu, Long Chen, Xing Liang, Jun-Feng Peng, Ming-Hui Zheng, Ju-Dong Li, Yong-Bing Cao, Cheng-Hao Shao, Identifying Liver Metastasis-Related Genes Through a Coexpression Network to Construct a 5-Gene Model for Predicting Pancreatic Ductal Adenocarcinoma Patient Prognosis, 2023, 52, 1536-4828, e151, 10.1097/MPA.0000000000002229
    3. Yuting Jiang, Chengyu Liao, Jianlin Lai, Yunyi Peng, Qilin Chen, Xiaoling Zheng, KRT7 promotes pancreatic cancer metastasis by remodeling the extracellular matrix niche through FGF2-fibroblast crosstalk, 2025, 15, 2045-2322, 10.1038/s41598-024-84129-1
  • Reader Comments
  • © 2021 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(3556) PDF downloads(185) Cited by(21)

Figures and Tables

Figures(13)  /  Tables(16)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog