Automated Image Segmentation and Evaluation Framework for Biomedical Applications

This project presents a robust framework for segmenting and evaluating biomedical images, focusing on vessel or region detection in grayscale imagery. The pipeline includes preprocessing steps like contrast enhancement, noise reduction, and adaptive thresholding, followed by morphological operations to improve segmentation accuracy. The segmented images are evaluated using performance metrics such as precision, recall, IoU (Intersection over Union), and error rate against ground-truth data. The framework provides flexibility for parameter tuning and can be adapted to various datasets with minimal adjustments, making it ideal for medical image analysis, such as retina vessel segmentation or tissue boundary detection.