A Novel Approach for K-means Optimization in Image Segmentation

One important approach for segmenting images is the K-means clustering algorithm. It is utilized in a variety of color spaces, including LAB and RGB. The extensive calculations are one of the k-means' drawbacks. Choosing the perfect k is another challenging task. In this article, we are targeting the first issue, which is the laborious calculations required to identify clusters. The computation is scaled down thousands of times. We achieved this by using DMC colors and an intermediate, limited-colors color space. In the K-means technique, we used the CIELab color space. The distance is calculated using Delta E (CIE 2000). Visually, the results were superior to those of other techniques.