idirectships/splat-amet-001

Amethyst Specimen (amet-001) — Gaussian Splat

Amethyst specimen reconstructed using Gaussian Splatting via COLMAP + opensplat pipeline on MacBook Pro.

Capture Method

• Device: Camera/iPhone (handheld)
• Total frames: 103 (images_jpg/ directory)
• Background: Dark matte (controlled background for COLMAP registration)
• COLMAP registration: Registered via colmap_ws_v1

Directory Structure

images_nobg/     — background-removed images (primary COLMAP input, symlinked as images/)
images_jpg/      — original JPG captures (103 frames)
images_cropped/  — cropped variants
images_heic/     — original HEIC captures
colmap_db/       — COLMAP database
colmap_ws/       — COLMAP workspace
colmap_ws_v1/    — COLMAP workspace v1 (sparse/ symlinked)
runs/
  v1-macbook-15k/
    output.ply   — 15,000 iteration Gaussian Splat
    cameras.json
    params.json
  2026-03-20-15000-Drus-MacBook-Pro/
    output.ply   — 15,000 iteration Gaussian Splat (with renders)
    renders/     — render outputs
    cameras.json
    params.json
    scores.json
  2026-03-20-30000-Drus-MacBook-Pro/   — 30,000 iteration run (params.json only; PLY not generated)
    renders/
    params.json
  macbook-15k/   — additional 15k run directory
pipeline.log
assess.log
sips-convert.log

PLY Iterations

Run	Iterations	Node	Notes
v1-macbook-15k	15,000	MacBook Pro	Standard run
2026-03-20-15000	15,000	MacBook Pro	Run with render outputs and scores
2026-03-20-30000	30,000	MacBook Pro	30k run (PLY incomplete)

Training Pipeline

1. Structure from Motion: COLMAP (feature extraction + matching + mapping)
2. Background removal: Applied before COLMAP input
3. Gaussian Splat training: opensplat
4. Output: .ply point cloud files viewable in SuperSplat

Viewing PLY Files

Use SuperSplat to view .ply Gaussian Splat files in-browser.