An Analysis of Image Segmentation Methods

Abstract

Image segmentation is a crucial stage in image processing because it allows for meaningful image analysis and information extraction. Image segmentation may be done in a variety of ways. This study examines a few of these approaches for grayscale images. MATLAB 8.1 was used to construct the paper. In order to analyze and extract information from these pictures, segmentation is essential. The segmentation of images has a promising future and has been the focus of modern study. The usual segmentation methods were addressed here: edge detection, clustering, and region growth. The problem with segmentation is that no one approach is suited for a certain type of picture, nor are all techniques applicable to all images.

Image Segmentation and Grouping Using Graph Method (SLIC Algorithm)

“…Several studies have explored the concept of image segmentation in recent years (Goyal, 2022;Sheeba et al, 2023;Mahmoud et al, 2018;Spoiala, 2023;Lakshmi & Kumar, 2022;Li et al, 2022;Kavitha et al, 2023). Image segmentation refers to classifying an image into different groups based on characteristics like intensity, texture, class or color, according to Goyal (2022).…”

“…Goyal (2022) notes that image segmentation is important for computer vision and industrial automation, as it improves accuracy and efficiency. It is used for tasks like object detection, extracting relevant parts from images (Spoiala, 2023;Lakshmi & Kumar, 2022;Li et al, 2022) and compressing images (Sheeba et al, 2023). Kavitha et al (2023) highlight its common use in healthcare for finding abnormalities and monitoring treatment progress.…”

“…However, image segmentation also presents challenges. Lakshmi and Kumar (2022) point out that no single method is suitable for all image types, and not all methods can be applied to every image. Francis et al (2022) further note the difficulty, due to factors like complex visual patterns, poor image quality, and similar-looking foreground and background elements.…”

“…However, image segmentation faces many challenges such as lack of universally applicable approaches, presence of noise, complex backgrounds and similar textures in foreground and background (Francis, Baburaj & George, 2022;Lakshmi & Anil Kumar, 2022). Unsupervised image segmentation which generates semantic segments without prior knowledge is also challenging (Jiao, 2022).…”