Note
Note records textual annotations extracted from the drawing. It stores the note text as a label, a note type distinguishing canvas-level notes from notes attached to a specific view, and a reference id. Notes frequently contain manufacturing instructions, inspection requirements, or references to standards, so capturing them accurately is vital for automation. The model includes a confidence score, allowing you to filter or route notes for review. Downstream, you can search, categorize, or translate notes programmatically without manually reading each sheet. The structured approach also supports machine learning models that classify note intent or detect risky instructions.
Bases: Feature
Represents a note in a technical drawing.
| PARAMETER | DESCRIPTION |
|---|---|
reference_id | Reference ID to identify the object. TYPE: |
label | A short description of the feature in a human-readable format. TYPE: |
confidence | Confidence in the feature extraction or interpretation. TYPE: |
note_type | The type of the note, specifying whether it is a canvas note, sectional note, or a sectional feature. TYPE: |