A Comparative Analysis of Otsu Thresholding and K-means Algorithm for Image Segmentation

Abstract

Image segmentation is an aspect of image processing and is used to find out the object in the image and dividing the image into different segments and discrete regions. The goal of image segmentation is to change the representation of an image into something that is more meaningful and easier to analyze and usually serves as the pre-processing before pattern recognition, feature extraction, and compression of the image. There are several challenges emerged in the field of image segmentation such as to differentiate between regions that may be determined by different factors, such as color, gray level or texture, overlapping objects can be difficult to separate also, shadowing can create additional borders. Many kinds of research have been done in the area of image segmentation. This research evaluates the two image segmentation algorithms Otsu thresholding and K-means using a parameter: Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Structural Content (SC) and Structural Similarity Index (SSIM) to measure the quality of segmented image. The lower value of MSE shows that the higher quality of segmented image is obtained. The higher value of PSNR and SSIM shows that for K-means original and segmented images do not loss the much of the property and have the higher similarities in input and segmented image. The lower SC value of an input and segmented image indicate the better quality of the image is generated by Otsu thresholding.

Keywords: Image Segmentation, MSE, PSNR, SSIM, SC, K-means, Otsu thresholding,