K-means Clustering for Image Segmentation Problem

(2) (6 marks) The array in Figure 3 shows feature vectors for each pixel in a 3 by 3 image. Apply the K-means clustering algorithm to assign each pixel to a region, what segmentation result will we get? Assume that (i) the method used to assess similarity is the Sum of Absolute Differences (SAD); (ii) there are two clusters, and the cluster centres are initially positioned at these positions in feature space: (5, 10, 15) and (10, 10, 25). Show the steps of how you get the answer. (5,5,5) (5,5,20) (5,10,25) (5,5,15) (20,5,15) (5,10,30) (5,10,15) (30,20,10) (20,30,10) Figure 3: The array of the features vectors. (2) (6 marks) The array in Figure 3 shows feature vectors for each pixel in a 3 by 3 image. Apply the K-means clustering algorithm to assign each pixel to a region, what segmentation result will we get? Assume that (i) the method used to assess similarity is the Sum of Absolute Differences (SAD); (ii) there are two clusters, and the cluster centres are initially positioned at these positions in feature space: (5, 10, 15) and (10, 10, 25). Show the steps of how you get the answer. (5,5,5) (5,5,20) (5,10,25) (5,5,15) (20,5,15) (5,10,30) (5,10,15) (30,20,10) (20,30,10) Figure 3: The array of the features vectors