Abstract
Conventional fuzzy clustering algorithms present several disadvantages with respect to image segmentation, including a tendency to arrive at local optima and a relatively high sensitivity to noise and initial cluster centers. To address these issues, we herein propose a kernel-based intuitionistic fuzzy clustering approach combining an improved grey wolf optimizer with a kernel-based intuitionistic fuzzy C-means clustering (IGWO-KIFCM) algorithm capable of carrying out differential mutations for image segmentation. The proposed method extracts spatial information from images and then applies a kernel-based intuitionistic fuzzy clustering objective function to improve the robustness of the algorithm against noise. To cope with the initial sensitivity and local optima issues, we develop an improved grey wolf optimizer based on differential mutation for the global optimization of the cluster centers. A comparative optimization assessment using six classic functions reveals that the improved grey wolf optimizer algorithm outperforms both the grey wolf optimizer and mean grey wolf optimizer algorithms in terms of searching ability and does not easily run into local optima. Moreover, the IGWO-KIFCM algorithm surpasses several other algorithms with respect to clustering performance across multiple datasets, and achieves good results in segmenting images with various types of noises.
Highlights
The fuzzy C-means (FCM) clustering algorithm is an unsupervised clustering analyzer based on the use of fuzzy sets
To address the disadvantages of the kernel-based IFCM clustering algorithm (KIFCM) algorithm in terms of sensitivity to initial cluster centers and the tendency to run into local optima, we drew inspiration from [9] in developing a strategy to update wolf pack locations based on dynamic random differential mutation for the optimization of cluster centers
To overcome the disadvantages of conventional intuitionistic FCM clustering algorithm (IFCM) algorithms in terms of sensitivity to noise and initial cluster centers, in this study we developed a kernel-based intuitionistic fuzzy clustering image segmentation algorithm using an improved grey wolf optimizer with differential mutation (IGWO-KIFCM)
Summary
The fuzzy C-means (FCM) clustering algorithm is an unsupervised clustering analyzer based on the use of fuzzy sets. To address the disadvantages of the KIFCM algorithm in terms of sensitivity to initial cluster centers and the tendency to run into local optima, we drew inspiration from [9] in developing a strategy to update wolf pack locations based on dynamic random differential mutation for the optimization of cluster centers By applying these methods, the proposed algorithm is expected to expand the global optimization range, improve local optimization precision, increase the probability of the algorithm jumping out of the local extreme value, and carry out more accurate image segmentation. As there will inevitably be pixels whose initial greyscale values are either 0 or 255, to ensure that authentic image information is retained, it is useful to develop strategies that guarantee that these pixels are not removed by the filter algorithm With this goal in mind, we propose a new method to extract spatial information from images to overcome the effect of noise on image segmentation.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
