Drone Floor Plan Annotator

Desktop Tkinter app to annotate windows on a floor-plan image and compute drone capture points, heading, and optional georeferenced outputs.

• Load floor-plan images (png, jpg, jpeg, bmp, tif, tiff, webp)
• Set plan origin
• Calibrate X and Y scale from known real distances
• Add window annotations with 3 clicks:
  1. window endpoint 1
  2. window endpoint 2
  3. side selector click (which side the drone should stand off)
• Two drone distance modes:
  • auto: computed from HFOV + fill ratio
  • manual: fixed user-entered distance (Manual dist m)
• Real-time update of existing annotations when you click Apply Settings
• Save/load project JSON
• Export CSV and verbose JSON
• Export marked plan image with drone positions overlaid
• View drone points on map from project JSON (project_map_viewer.py)

After origin is set:

• image x: positive to the right
• image y: positive downward
• local east: positive to the right
• local north: positive upward

Conversion:

• east_m = (px_x - origin_x) * meters_per_pixel
• north_m = -(px_y - origin_y) * meters_per_pixel

Formula:

d_m = W_m / (2 * tan(HFOV / 2) * fill_ratio)

Where:

• W_m is window width in meters
• HFOV is horizontal field of view in degrees
• fill_ratio is desired frame occupancy ratio (0 < ratio <= 1)

• Set Dist mode to manual
• Enter Manual dist m
• Click Apply Settings
• Manual distance overrides auto formula (HFOV/fill are ignored for standoff)

When you click Apply Settings, the app recalculates existing annotations immediately (no reload needed) using the current settings and distance mode.

• New annotations are sequentially numbered (W1, W2, ...)
• Preview rendering does not consume IDs
• IDs are re-numbered sequentially when loading a project and after deletions

If georef fields are set:

• origin latitude/longitude/altitude
• plan +Y azimuth clockwise from true north

the app converts local EN coordinates to LLA using navpy.

• Saves editable project state
• Saved under output/project/json
• Includes annotation geometry and current settings

• Verbose structure with:
  • project (full project payload)
  • exports (flat per-annotation rows)

• Flat per-annotation table for spreadsheets/scripting

• Exports plan image with drone positions, guide lines, and labels drawn
• Menu: File -> Export Marked Plan Image...

• Select: select annotation
• Pan: tool mode (mouse pan is still middle/right drag)
• Origin: set origin point
• Cal X: set X calibration
• Cal Y: set Y calibration
• Window: add window annotation

• Mouse wheel: zoom
• Middle drag or right drag: pan
• Left click: active tool action

• Esc: switch to Select and clear temp tool state
• Ctrl+S: save project
• Ctrl+Z: undo
• Ctrl+Y: redo

• output/project/json: project saves + JSON exports
• output/project/csv: CSV exports
• output/project/images: marked plan image exports

exporter.py also writes output CSV into output/project/csv.

project_map_viewer.py loads project JSON and plots drone points on a map.

Run:

python project_map_viewer.py

Or with file:

python project_map_viewer.py path/to/project.json

exporter.py converts saved/exported JSON to a compact CSV with:

• id
• drone_lat_deg
• drone_lon_deg
• drone_alt_m
• heading_deg

Run:

python exporter.py

Or:

python exporter.py path/to/input.json output_name.csv

• Python 3.10+
• Tkinter
• Dependencies in requirements.txt

Install:

pip install -r requirements.txt

python app.py