N-Views egg geometry dataset

This dataset contains egg geometries, including manual measurements and photography (3D photogrammetry, monocular silhouette and N-Views methods) datasets for 6 samples of Coturnix coturnix (CC) eggs, 6 samples of Gallus gallus domesticus (GG) eggs, and microscopic digital images of two samples of the predatory mite species Amblyseius swirskii (AS).

The data is designed to validate and benchmark photography-based measurement techniques for the reconstruction of ovoid structures.

Dataset structure

1. ImageryEggsDataset.zip

The primary imagery archive, which has the following internal structure:

./0_Photogrammetry/

Benchmarking schema for 12 egg samples reconstructed via photogrammetry (structure-from-motion).

• • /R_CC1/ to /R_GG6/: 12 folders each containing:
    • /raw_images/: 100-150 raw images (.JPG, 3456×2304 px) of an egg geometry using a turntable and fiduciary markers.
    • R_CC1_b.mg: Meshroom (Alicevision 2021) node graph for the structure-from-motion reconstruction pipeline.
    • R_CC1_b.stl: Scaled 3D output model of the egg geometry.
    • texture_1001.png, texturedMesh.mtl, texturedMesh.obj: Textured 3D reconstruction files.
  • MarkerU.pdf: Fiduciary marker used for scale/rotation.
  • Readme.txt: Printed marker coordinates.

./1_SilhoutteBestEffort/

Monocular top-view images for Pappus theorem reconstruction (70 mm focal length).

• • /calibration/: Checkerboard images (9×6, square size: 25 mm). 
  • /images_h/: 12 top-view images of CC and GG samples.
  • /masks/: Precomputed egg contours.
  • /out_annots/: Annotations of contours, main axis, and side lobes.
  • pappus_results.csv: Summary statistics for the Silhouette approach.
  • triad_arucomarker.pdf: Fiduciary markers in the scheme consist of the Aruco ids [0,1,2] with printed size 22 mm and separation of 7.5 mm.

./2_NViews/

Multi-view reconstruction approach for the 12 CC and GG egg samples.

• • /R_CC1/ to /R_GG6/: Individual sample folders containing:
    • /calibration/: Checkerboard images (9×6, square size: 25 mm).
    • /images_h/: 8 multi-view photographs.
    • /masks/: Precomputed egg contours.
    • calibration_outoput.json: Distortion and camera intrinsic matrix (OpenCV 5-parameter distortion model).
    • result.json
      

      📤 Share this page

      Found this useful? Share it with your network.

      𝕏XfFacebookinLinkedIn✆WhatsApp✈Telegram✉EmailMMendeley⧉Copy LinkR^GResearchGateAAcademia.edu

      ✓ Link copied! Paste it on ResearchGate / Academia.edu