Lefika la Noka Tufa: Paired PPL–XPL Scans and Photomicrographs
Description
This dataset contains multiscale petrographic images of the Lefika la Noka freshwater tufa deposit in the Cradle of Humankind, Gauteng Province, South Africa. The collection was produced to support sedimentological, petrographic and palaeoenvironmental investigations of the deposit and its range of primary, reworked and diagenetically modified tufa fabrics. The dataset represents 52 thin sections and contains 104 whole-thin-section flatbed scans, comprising paired plane-polarized light (PPL) and cross-polarized light (XPL) images. It also contains 5,567 optical microscopy images, including 2,747 verified matched PPL/XPL photomicrograph pairs, section-overview images and a small number of auxiliary or unpaired images. The microscope images are generally 1,024 × 768 pixels and include 500 µm scale bars. Files are organized by thin-section identifier, microscope-field identifier and illumination mode. The whole-section scans preserve the larger-scale distribution and relationships of the tufa fabrics, whereas the photomicrographs provide higher-resolution information about lithoclasts, extraclasts, intraclasts, micritic matrix, primary depositional framework, crystalline cement and pore space. Together, the two image scales allow petrographic observations to be interpreted within their broader sedimentological context. The dataset can support tufa lithofacies characterization, petrographic point counting, quantitative image analysis, comparison of PPL and XPL optical properties, assessment of sedimentary disturbance and recovery, and the development or testing of computer-vision and machine-learning methods. Potential applications include automated image classification, segmentation, petrographic similarity searches, point-counting assistance and cross-scale localization of microscope fields on whole-thin-section scans. The PPL and XPL photomicrographs are already paired and aligned with one another. However, the microscope fields were collected independently of the flatbed scans and are not presently registered to precise coordinates on the whole-section images.
Files
Steps to reproduce
The dataset can be reproduced by preparing standard petrographic thin sections from geological samples and assigning each section a unique identifier. Digitize the complete thin sections using a flatbed scanner under consistent lighting, orientation, resolution and image-format settings. Examine the thin sections with a polarizing microscope and photograph representative fields of view in both plane-polarized light (PPL) and cross-polarized light (XPL). Capture each PPL–XPL pair without moving the specimen so that the two images remain aligned. Record the magnification or include a calibrated scale bar. Retain the images at their original resolution, apply a consistent file-naming system and organize them according to thin-section identifier and illumination mode. Finally, verify image quality, PPL–XPL pairing, scale information, filenames and folder assignments before compiling the files into the published dataset. Any image processing or derived measurements should be documented and stored separately from the original images.
Institutions
- University of JohannesburgGauteng, Johannesburg