INDUSTRY FIELD GUIDE
Cabinet door and drawer reveal data for AI inspection
An installed door can look misaligned because the camera is tilted.
NAICS 337110 · Wood Kitchen Cabinet and Countertop Manufacturing
NAICS 337212 · Custom Architectural Woodwork and Millwork Manufacturing
An installed door can look misaligned because the camera is tilted. A training set can show the difference between a visual impression and a measured reveal. The Woodwork Institute runs a shop drawing inspection program. micro1’s workplace video listing names cabinetmaking and woodwork among its settings. A static quality-inspection dataset and a first-person recording project are separate offers with separate requirements. Woodwork Institute shop drawing inspection · micro1 workplace recording listing
On an owned demo cabinet, compare an even gap shown from an oblique angle with a visibly uneven gap confirmed by a gauge. Put the measured checkpoints in a simple elevation sketch. Add a drawer-front example with another finish to see whether the model overuses contrast. The reviewer should specify whether the issue is gap consistency, alignment, or evidence too weak to judge. Use neutral acceptance rules created for the exercise.
Show the gauge and the defect without including a client’s drawings, address, or proprietary hardware detail. Confirm who owns a finished cabinet design before photographing production work.
Offer a demo unit or inspection sample
Original Fiund exercise
A dark edge is not necessarily a wider reveal
An original paper cabinet mockup is captured with front light and side light. A drawn scale supports a toy comparison.
| Record | Source | What it records |
|---|---|---|
| 1 | Reference view | Left and right reveal targets 3 mm; no acceptance tolerance given. |
| 2 | Side-lit image | Right edge appears dark and broad, but ruler shows 3 mm on both sides. |
| 3 | Front-lit image | Dark region disappears; scaled gaps remain 3 mm on both sides. |
Question for the model
What should the model label about the right reveal?
Show the reference answer and checks
Both measured reveals are 3 mm in records 2 and 3, matching the reference value. The extra dark width changes with lighting and should be labeled as a visual artifact rather than measured gap asymmetry. The example does not specify an acceptance tolerance.
- Using shadow width as a physical dimension
- Treating one image artifact as proof of misalignment
Fields to preserve
- lighting_view
- Documents capture conditions
- calibrated_gap
- Separates physical measurement from appearance
- side_id
- Aligns left and right observations
Test another case
Give unequal scaled gaps under both lighting setups; score real asymmetry separately from shadow.
Use owned mockups; no actual cabinet drawings, customer homes, installation or quality grading.
Discuss a contribution
Describe the sources you control, the task they could support and what has already been checked. Use the existing-data assessment for an archive or the collection brief for a new recording project. The contributor agreement defines Fiund’s licensing authority and compensation. A guide or fictional exercise does not establish an accepted collection or a sale.