I’m working on a electrical engineering project and need an explanation to help me understand better.
K-Means Clustering Based Grayscale Image
Segmentation
Grayscale images carries only intensity information of the image. K-means clustering clustering
based grayscale image segmentation is a clustering algorithm that groups image pixels into k
groups based on their intensity characteristics. The grouping is done by minimizing the sum of
the distances between each pixel and the cluster centroid.
Design the matlab code to implement the K-Means clustering based grayscale cameraman.tif
image segmentation:
1. Initialization: choose the number of groups K. K centroids are established by randomly
choosing K pixels from the cameraman.tif image.
2. Assign each pixel to its nearest centroid.
3. The position of the centroid of each group is updated taking as the new centroid based
on the average position of the objects belonging to the same group.
4. Repeat steps 2 and 3 until the centroids move below a threshold.
In your project report, show your matlab design code, the chosen parameters of K, and the
threshold in the above steps.
Show the final picture of each segmentation cluster and the overall combined segmentation
result picture. I uploaded an example code. Need to follow and modify the given K-Mean code to finish this project.
0 comments