A Survey on Image Segmentation Techniques

Abstract

The essential step in digital image processing is segmentation which can be used to partition the images into particular regions or objects and the level of partitioning depends on the individuality of the problem being solved. Segmentation of the image is widely categorized into two. The first one is Discontinuity which measures the sudden changes of intensities for partitioning the image and the other category Similarity is measured, based on the predefined methods such as thresholding, region growing, splitting and merging. The images are considered as inputs for performing segmentation and the result is attributes extracted from the images. Segmenting an image is an initial step to understand and analyze what is inside the image and this will be done mandatorily for all medical imaging analysis. Several segmentation techniques have been proposed in the past, but none of the segmentation methods are invented without any drawbacks. Hence, this study discusses a review on the various segmentation techniques of image which will help in further advancement in this field.