Skip to Main Content
Several general-purpose algorithms and techniques have been developed for image segmentation. Since there is no general solution to the image segmentation problem, these techniques often have to be combined with domain knowledge in order to effectively solve an image segmentation problem for a problem domain. This paper presents a comparative study of the basic image segmentation techniques i.e. edge-based, k-means clustering, thresholding and region-based techniques, using a number of test images. The objects extracted after image enhancement and image segmentation as compared to the objects desired, whether the region boundaries are closed or disconnected and the mean weighted distance measure of the segmented objects with respect to the original image form the criteria to perform the comparative study. Image segmentation is further used for object matching between two images. Correlation between the objects being matched in the two images is used as a measure of similarity between the two objects. The first principal component axis, determined by principal component analysis (PCA), of the objects being matched are aligned with the x-axis to take into account the different orientation of an object in different images.