OCR Coordinates to Geometry

Русский · English

Turn a screenshot of a coordinate table into numbered points, a line and an optional polygon directly in QGIS.

OCR2Geometry in QGIS 4 with source CRS and output-layer naming.

Why this plugin exists

Mining licences, land-allocation documents and legacy survey reports often contain corner coordinates only as scanned tables. The traditional workflow is OCR → spreadsheet cleanup → CSV import → point sorting → polygon creation. This plugin reduces that workflow to a screenshot, a review table and one button.

Current features — v1.0

• Open PNG/JPG images or paste a screenshot from the clipboard.
• Offline recognition with RapidOCR.
• One-click OCR dependency installation inside the QGIS user profile.
• Editable seven-column DMS preview.
• Validation of degrees, minutes, seconds and point numbers.
• Numbered corner-point layer.
• Open or closed line in point-number order.
• Optional polygon layer.
• Source CRS selection from the standard QGIS catalog (EPSG:4326 by default).
• Custom base names for output layers.
• No image or coordinate upload to external services.
• English and Russian interface selected from the QGIS locale.
• OCR2Geometry icon, diagnostics and persistent user options.
• Automatic DMS, degrees/decimal-minutes and decimal-degree parsing.
• Latitude/longitude reversal, N/S/E/W hemispheres and signed coordinates.
• Plain-text coordinate paste and source-order controls.
• Live decimal-degree preview columns.
• Pre-creation coordinate and geometry quality review.
• Missing point-number, coincident-vertex and invalid-polygon warnings.
• CRS-aware line length, perimeter and area calculation.
• Green, yellow and red OCR-confidence highlighting with percentage tooltips.
• CSV export of the reviewed table with DMS, decimal coordinates and OCR confidence.
• Spreadsheet-style manual entry with Excel paste and live two-way DMS/DD conversion.
• Configurable second precision from 0 to 6 decimal places (3 by default).

Supported input layout:

Point | Lat deg | Lat min | Lat sec | Lon deg | Lon min | Lon sec

Install

1. Download ocr_coordinates_to_geometry.zip from the latest GitHub release.
2. In QGIS open Plugins → Manage and Install Plugins → Install from ZIP.
3. Select the downloaded ZIP and start the plugin.
4. On first recognition, allow the plugin to install RapidOCR. Internet access is needed only for this initial dependency download.

Compatibility: QGIS 4.0–4.99, Windows 10 and Windows 11.

Basic workflow

1. Start the plugin from Vector → OCR2Geometry → OCR2Geometry or its toolbar icon.
2. Open or paste a coordinate-table image, then click Recognize.
3. Review the recognized rows and choose the source CRS.
4. Enter a base name and select the required point, line and polygon outputs.
5. Click Add to QGIS project and verify the geometry against the source document.

The selected name is supplemented with separate points, line and polygon suffixes. The selected CRS is assigned to the layers; use standard QGIS tools when reprojection is required.

Stable release policy

Version 1.0 is stable and feature-frozen. Maintenance releases focus on confirmed defects and compatibility; larger ideas remain deferred until users show a concrete need. Coordinate transformations remain a standard QGIS task.

Accuracy and CRS warning

Always review recognized values before using the geometry and select the CRS that describes the source coordinates. Assigning a CRS does not transform the numbers. Use QGIS export/reprojection tools for coordinate transformations.

Development

python -m unittest discover -s tests -v
python scripts/build_zip.py

Bug reports, real-world sample layouts and pull requests are welcome. Please read CONTRIBUTING.md before sharing documents.

Project documents

• Roadmap
• User guide
• Troubleshooting
• Third-party notices
• Changelog
• Windows OCR setup
• Contributing
• Russian README

License

MIT. Optional OCR components are not bundled; see the third-party notices.

Developers

• Markseder
• Matveev Pavel — pavelmatveev84@gmail.com