Description:

This Master Prompt develops an evidence-based framework for separating meaningful structures, objects, phases and boundaries from digital images while preserving their physical and analytical significance. It addresses segmentation-problem definition, image preparation, reference annotations, thresholding, color-based clustering, edge detection, contour extraction, mathematical morphology, connected-component analysis, watershed segmentation, learned boundary prediction, object identification and validation. The framework connects the selection of a segmentation method with the imaging modality, available contrast, object geometry, spatial resolution, intended measurement and supporting reference evidence.