A survey on the utilization of Superpixel image for clustering based image segmentation

Area of Science:

• Computer Vision
• Image Processing
• Machine Learning

Background:

• Superpixels enhance image segmentation accuracy and reduce computational load, especially in noisy conditions.
• Clustering is a proven image segmentation technique with effective performance across various applications.
• The synergy between superpixel generation and clustering techniques is crucial for optimal image segmentation.

Purpose of the Study:

• To provide an up-to-date survey on superpixel and clustering techniques for image segmentation.
• To review existing literature on combining superpixel and clustering strategies.
• To comparatively evaluate the efficacy of superpixel combined with partitional clustering methods.

Main Methods:

• Overview of various superpixel image generation techniques.
• Detailed examination of clustering techniques, focusing on efficient partitional clustering, including challenges and solutions.
• Review of existing literature combining superpixel and clustering for diverse image segmentation tasks.
• Comparative analysis of superpixel and partitional clustering combinations on oral pathology and leaf images.

Main Results:

• The study provides a comprehensive understanding of superpixel generation strategies.
• It elucidates how different superpixel approaches impact partitional clustering performance.
• Evaluations demonstrate the efficacy of combined superpixel and partitional clustering methods in specific image datasets.

Conclusions:

• The combination of superpixel and partitional clustering offers a robust approach to image segmentation.
• Understanding the interplay between superpixel generation and clustering is key to selecting optimal methods.
• This survey offers valuable insights for researchers and practitioners in image segmentation.