What you can upload
ZeroCount currently accepts PDF drawings only. The upload zone restricts the file picker and any drag-and-drop to .pdf files — other types are silently rejected.
PDFs can be:
- Single-page or multi-page (each page becomes its own file inside the service).
- Vector or raster — both work. Vector PDFs give the cleanest detection, but raster scans of paper drawings are also handled.
- Any size — there's no hard page-size limit, but very large drawings (A0 at high DPI) take longer to upload and process.
What you cannot upload directly
| Format | Workaround |
|---|---|
| DWG / DXF (AutoCAD) | Export to PDF from your CAD package first. Most CAD tools have a Print to PDF or Export → PDF option. |
| Images (PNG, JPG, TIFF) | Drop them into a PDF — either via a print-to-PDF tool, or by inserting them into a blank Word/Pages document and exporting. |
| Revit, IFC, native BIM files | Export the relevant sheets to PDF from the BIM tool. |
Image and DWG support may come later. For now, the rule of thumb is: if it doesn't end in .pdf, it doesn't upload.
Preparing PDFs for the best result
The detection pipeline cares about a few things:
- Orientation. ZeroCount can read drawings in any orientation, but the OCR pass works better when text and the legend are upright. Rotate the file after upload if needed — see Rotating files.
- Legend clarity. A clean, complete legend with one symbol per row gives the cleanest auto-detection. Hand-marked legends, very small legends, or legends with overlapping symbols may need manual correction in Stage 2.
- Backgrounds. Heavy shading or background imagery can confuse template matching. If detection misses obvious symbols, try Remove Background — see Remove background.
- Resolution. Scanned drawings at 200 DPI or higher detect better than 72 DPI screen captures. If you control the export, use 300 DPI.
How files are stored
Once uploaded, a PDF is processed and cached on the server — you don't need to upload it again to come back to it. Files stay with their project for as long as the project exists; archiving a project keeps the files; deleting a file from the project removes it permanently.