Image Segmentation Methods: Overview, Challenges, and Future Directions

摘要

Image segmentation is a popular topic in computer vision which involves partitioning a digital image into multiple segments or regions. It aims to simplify or change the representation of an image into a new version which is more meaningful and easier to interpret. This technique is widely used in various applications, including medical imaging, autonomous vehicle tracking, object recognition, satellite image analysis, and industrial robotics. Fundamental image segmentation methods can be grouped into two categories: edge-based, and region-based methods. In this paper, we review the existing image segmentation methods, comparing their strengths and weaknesses. We also review few quantitative metrics used to evaluate the performance of image segmentation methods. Then, we use three metrics to evaluate the performance of segmentation methods on color and grayscale images. Additionally, we show a bibliometric analysis of publications on image segmentation by using 10,000 top cited scientific papers related to this area.