Color-Based Image Segmentation Using K-Means Clustering Approach
This study aims to develop and implement image processing functions using the histogram of a single, two-dimensional butterfly image. The specific objective is to classify the color feature sets of image pixels. The focus of the study is on color-based image segmentation, which is based on the assumption that regions of homogeneous colors within an image correspond to distinct clusters representing meaningful objects. The initial approach of the study involves the developing and implementing image processing functions based on the histogram of the butterfly image. These functions aim to extract essential color features from the image pixels and subsequently classify the image based on these features. To achieve this, the study employs a k-means clustering approach, resulting in a three-cluster solution: Cream Color Cluster, Yellow Color Cluster, and Orange Color Cluster. The analysis reveals significant dissimilarities among the three clusters obtained from the butterfly image in terms of histograms and visual features derived from the L*a*b* color space. Through this approach, valuable insights into the intricate color patterns exhibited by butterfly wings can be uncovered.
Keywords: Image segmentation, color-based, histogram, k-means clustering, MATLAB.
