Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation
As a classic clustering method, the traditional k-Means algorithm has been widely used in image processing and computer vision, pattern recognition and machine learning. It is known that the performance of the k-means clustering algorithm depends highly on initial cluster...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Thi-Qar
2019-05-01
|
| Series: | مجلة علوم ذي قار |
| Subjects: | |
| Online Access: | http://jsci.utq.edu.iq/index.php/main/article/view/216 |
| _version_ | 1797404236052430848 |
|---|---|
| author | Abbas H. Hassin Alasadi |
| author_facet | Abbas H. Hassin Alasadi |
| author_sort | Abbas H. Hassin Alasadi |
| collection | DOAJ |
| description |
As a classic clustering method, the traditional k-Means algorithm has been widely used in
image processing and computer vision, pattern recognition and machine learning. It is known that
the performance of the k-means clustering algorithm depends highly on initial cluster centers.
Generally initial cluster centers are selected randomly, so the algorithm could not lead to the
unique result. In this paper, we present a method to compute initial centers for
k-means clustering. Our method based on an efficient technique for estimating the modes of a
distribution. We apply the new method in segmentation phase of color images. The experimental
results appeared quite satisfactory.
|
| first_indexed | 2024-03-09T02:52:01Z |
| format | Article |
| id | doaj.art-3497e5d138a444d59617f3dc0bf31073 |
| institution | Directory Open Access Journal |
| issn | 1991-8690 2709-0256 |
| language | English |
| last_indexed | 2024-03-09T02:52:01Z |
| publishDate | 2019-05-01 |
| publisher | University of Thi-Qar |
| record_format | Article |
| series | مجلة علوم ذي قار |
| spelling | doaj.art-3497e5d138a444d59617f3dc0bf310732023-12-05T09:57:42ZengUniversity of Thi-Qarمجلة علوم ذي قار1991-86902709-02562019-05-0131Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation Abbas H. Hassin Alasadi0Basrah University As a classic clustering method, the traditional k-Means algorithm has been widely used in image processing and computer vision, pattern recognition and machine learning. It is known that the performance of the k-means clustering algorithm depends highly on initial cluster centers. Generally initial cluster centers are selected randomly, so the algorithm could not lead to the unique result. In this paper, we present a method to compute initial centers for k-means clustering. Our method based on an efficient technique for estimating the modes of a distribution. We apply the new method in segmentation phase of color images. The experimental results appeared quite satisfactory. http://jsci.utq.edu.iq/index.php/main/article/view/216k-Means algorithmImage segmentationColor spaces. |
| spellingShingle | Abbas H. Hassin Alasadi Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation مجلة علوم ذي قار k-Means algorithm Image segmentation Color spaces. |
| title | Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation |
| title_full | Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation |
| title_fullStr | Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation |
| title_full_unstemmed | Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation |
| title_short | Applying New Method for Computing Initial Centers of k-Means Clustering with Color Image Segmentation |
| title_sort | applying new method for computing initial centers of k means clustering with color image segmentation |
| topic | k-Means algorithm Image segmentation Color spaces. |
| url | http://jsci.utq.edu.iq/index.php/main/article/view/216 |
| work_keys_str_mv | AT abbashhassinalasadi applyingnewmethodforcomputinginitialcentersofkmeansclusteringwithcolorimagesegmentation |