Image Segmentation Based on Statistical Confidence Intervals

Area of Science:

• Computer Vision
• Image Processing
• Statistical Modeling

Background:

• Image segmentation is crucial for image analysis, partitioning images into meaningful regions.
• Existing segmentation methods face challenges with noise, particularly speckle noise.
• The Otsu algorithm is a widely used thresholding technique for image segmentation.

Purpose of the Study:

• To develop a new image segmentation method robust to speckle noise.
• To enhance image analysis by improving segmentation accuracy in noisy conditions.
• To compare the proposed method against the standard Otsu algorithm.

Main Methods:

• A novel image segmentation technique integrating statistical confidence intervals with the Otsu algorithm.
• Numerical experiments were conducted to evaluate the method's performance.
• Validation was performed using various image samples, including those with speckle noise.

Main Results:

• The proposed method demonstrates superior performance compared to the standard Otsu algorithm for images with speckle noise.
• The algorithm effectively mitigates the impact of speckle noise entropy.
• Experimental results validate the efficacy of the new segmentation approach.

Conclusions:

• The developed image segmentation method offers improved robustness against speckle noise.
• This technique provides a more effective solution for segmenting noisy images.
• The integration of statistical confidence intervals enhances the Otsu algorithm's capabilities.