H.5.7. Segmentation
Mohsen Erfani Haji Pour
Abstract
The segmentation of noisy images remains one of the primary challenges in image processing. Traditional fuzzy clustering algorithms often exhibit poor performance in the presence of high-density noise due to insufficient consideration of spatial features. In this paper, a novel approach is proposed that ...
Read More
The segmentation of noisy images remains one of the primary challenges in image processing. Traditional fuzzy clustering algorithms often exhibit poor performance in the presence of high-density noise due to insufficient consideration of spatial features. In this paper, a novel approach is proposed that leverages both local and non-local spatial information, utilizing a Gaussian kernel to counteract high-density noise. This method enhances the algorithm's sensitivity to spatial relationships between pixels, thereby reducing the impact of noise. Additionally, a C+ means initialization approach is introduced to improve performance and reduce sensitivity to initial conditions, along with an automatic smoothing parameter tuning method. The evaluation results, based on the criteria of fuzzy assignment coefficient, fuzzy segmentation entropy, and segmentation accuracy, demonstrate a significant improvement in the performance of the proposed method.
H.6.4. Clustering
P. Shahsamandi Esfahani; A. Saghaei
Abstract
Data clustering is one of the most important areas of research in data mining and knowledge discovery. Recent research in this area has shown that the best clustering results can be achieved using multi-objective methods. In other words, assuming more than one criterion as objective functions for clustering ...
Read More
Data clustering is one of the most important areas of research in data mining and knowledge discovery. Recent research in this area has shown that the best clustering results can be achieved using multi-objective methods. In other words, assuming more than one criterion as objective functions for clustering data can measurably increase the quality of clustering. In this study, a model with two contradictory objective functions based on maximum data compactness in clusters (the degree of proximity of data) and maximum cluster separation (the degree of remoteness of clusters’ centers) is proposed. In order to solve this model, a recently proposed optimization method, the Multi-objective Improved Teaching Learning Based Optimization (MOITLBO) algorithm, is used. This algorithm is tested on several datasets and its clusters are compared with the results of some single-objective algorithms. Furthermore, with respect to noise, the comparison of the performance of the proposed model with another multi-objective model shows that it is robust to noisy data sets and thus can be efficiently used for multi-objective fuzzy clustering.
