Textile Microtexture Dataset: 100+ Annotated Images & 40+ Quantitative Classification Features

This dataset comprises high-resolution images of diverse textile fabrics alongside a comprehensive CSV file of microtexture features extracted for machine learning analysis. Each image is paired with quantitative descriptors, including Haralick, LBP, fractal, wavelet, and edge-based metrics. This enables detailed classification and defect detection tasks. The dataset includes multiple fabric types (ex., cotton, wool, polyester, silk) and covers a range of microtexture patterns, with all images captured using a custom imaging rig. Custom imaging rig consists of Raspberry Pi High Quality Camera with a 100X Lens. Feature extraction methods and parameters are documented for reproducibility. This resource supports research in computer vision, textile engineering, and automated quality assessment.