Python-Based Image Processing for Projected Void Quantification from SEM Fracture Surfaces

This dataset contains the Python image-processing code used for projected void quantification from SEM fracture-surface images. The workflow segments dark projected cavities, applies morphological filtering, separates clustered voids using distance-transform/watershed segmentation, and extracts quantitative descriptors including projected void area, equivalent diameter, aspect ratio, fill ratio, void count, void number density, and projected void area fraction. The routine also generates colored overlay images for visual validation and exports individual and summary void measurements as CSV files. The code was developed to support reproducible quantitative fractography of impact-fractured additively manufactured polymer-composite specimens.本数据集包含用于从扫描电子显微镜（SEM）断口图像中量化投影孔隙的Python图像处理代码。该工作流程首先对暗部投影空腔进行分割，应用形态学滤波，再通过距离变换/分水岭分割算法分离团聚孔隙，并提取多种量化描述符，包括投影孔隙面积、等效直径、长宽比、填充率、孔隙总数、孔隙数密度以及投影孔隙面积占比。该程序还可生成彩色叠加图像用于视觉验证，并将单孔隙测量结果与汇总孔隙测量数据导出为逗号分隔值（CSV）文件。本代码的开发旨在为冲击断裂的增材制造聚合物复合材料试样提供可复现的定量断口分析支持。