An Efficient Workflow for Background Masking in Small-Object Photogrammetry
Bibliographic record
Abstract
Modern photogrammetry software offers a relatively inexpensive and very high quality technique to produce textured 3D models of museum objects. The efficacy of this technique has already been demonstrated by Queen’s students in digitizing Greco-Roman vessels from the Department of Classics’ Diniacopoulos Collection, terra-cotta figurines from National Institution Stobi in the Republic of Macedonia, and wood fragments from The White Coffin held in the Art Conservation Program at Queen’s. A significant bottleneck, however, in the workflow remains the removal of the background from each of the input photos when using a rotary table and tripod-mounted camera; with the object moving on the rotary and the background remaining fixed, the software becomes confused in its reconstruction of the scene and no full-3D model can be created. The process to remove the background by hand can take upwards of 10 hours for a single object. Proposed is a new, efficient workflow that uses intelligent edge-detection algorithms to automatically separate in-focus from out-focus areas in the scene. With only a small modification to the photographic parameters to reduce depth-of-field, the out-of-focus background can be removed for large datasets (>150 images) in less than 30 minutes with minimal human intervention. Further, recently released AI-based subject selection algorithms as part Adobe Photoshop promise to speed this process even further and to radically simplify the use of photogrammetry in creating interactive virtual exhibits, and 3D-printable data. Much work still needs to be done in evaluating the robustness of this AI "black-box" from Adobe in photogrammetry applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.015 | 0.018 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".