J1

Severely curious. Plant Breeding, Quant. Gen., Data Science, Remote Sensing, A.I., and a bunch of random interests.

4 minute read

Overview

This project consists of two tightly related Tkinter desktop GUIs for working with large GeoTIFFs, particularly drone-derived orthomosaics. One tool is a GeoTIFF grid builder and labeling application focused on structure, geometry, and reliable export of spatial data, while the other is a lightweight orthomosaic viewer and cropper optimized for navigation, inspection, and fast extraction of native-resolution subsets. They are intentionally kept separate to avoid overloading a single interface, and together they cover the manual geospatial tasks I found myself repeating most often.

Tool 1, the GeoTIFF Grid Builder, is designed for structure and export: it loads and preprocesses GeoTIFFs, displays them using a multi-resolution image pyramid, draws real-world-sized and rotated grids, labels grid cells using serpentine logic, and exports the resulting geometries as ESRI Shapefiles. Tool 2, the Orthomosaic Viewer and Cropper, is optimized for navigation and extraction: it enables smooth panning and zooming of massive rasters, displays both pixel and map coordinates, and exports the current viewport as a native-resolution GeoTIFF while preserving the original CRS, geotransform, and nodata values. Both tools share the same design philosophy—native GDAL reads and writes, minimal GIS dependencies, desktop-first mouse-driven interaction, and explicit safeguards around coordinate reference systems and pixel geometry.

Animated GIF showing grid workflow

Animated GIF showing cropper workflow

Tool 1: GeoTIFF Grid Builder

What this tool is

A Tkinter desktop application for loading GeoTIFF imagery, preprocessing it into a clean, validated TIFF, displaying it efficiently, drawing configurable grids, labeling cells, and exporting those grids as shapefiles.

It is meant for:

  • Plot grids
  • Experimental layouts
  • Structured sampling
  • Any situation where grid geometry must match real-world dimensions

Primary Workflow (End-to-End)

  1. Launch

    • Run main.py
    • A Tkinter window opens and instantiates GeoTIFFApp from app.py

  2. Open & Pre-process TIFF

    • Click “Open & Pre-process TIFF”

    • A guided wizard runs in the background:

      • Reads the input GeoTIFF
      • Optionally strips alpha bands
      • Optionally downsamples large images
      • Writes a cleaned _clean.tif
      • Builds a multi-level PIL image pyramid
      • Runs a full block-by-block integrity check
      • Displays a preview
      • Enables “Load in Viewer” only if validation passes

  3. Load into Viewer

    • Spatial metadata is read:

      • CRS
      • Units
      • Geotransform
    • Projection details appear in the sidebar
    • Warnings are shown for geographic CRS (degrees)
    • The image pyramid is rendered to the canvas

  4. Grid Creation

    • Double-click to set the grid anchor point

    • Configure:

      • Rows / columns
      • Cell size
      • Spacing
      • Rotation
    • Draw the grid on the canvas

  5. Labeling & Export

    • Generate serpentine numbering based on:

      • Start corner
      • Direction
      • Prefix

    • Export grid polygons as a shapefile with attributes

UI Layout & Controls

Main Window (widgets.py)

  • Fixed-width left sidebar with collapsible sections
  • Right-side canvas with scrollbars
  • Bottom status bar for warnings and messages

Menus

  • TIFF

    • Launch preprocessing wizard
    • Pyramid level slider

  • Grid

    • Draw / reset grid
    • Grid parameter controls

  • Numbering

    • Start corner & direction
    • Start number and prefix
    • Generate / reset numbering
    • Export shapefile

Canvas Interactions

  • Double-click: set anchor
  • Arrow keys: nudge anchor
  • Mouse wheel: change pyramid level

Core Modules

  • main.py — Minimal entry point
  • app.py — App bootstrap and wiring
  • wizard.py — Background preprocessing workflow + UI logging
  • widgets.py — Tkinter layout and menu wiring

Services & Responsibilities

  • preprocess.py

    • Reads GeoTIFF via GDAL
    • Strips alpha bands
    • Downsamples if needed (default ~100 MB target)
    • Writes cleaned TIFF
    • Generates display pyramid

  • check_tiff_integrity.py

    • Reads every block of every band
    • Ensures the output TIFF is fully readable

  • metadata.py

    • Parses CRS and geotransform
    • Computes pixel sizes (meters where possible)
    • Handles pixel ↔ geographic coordinate conversion

  • image.py

    • Manages pyramid levels
    • Maintains scale factors
    • Renders the active display image

  • grid.py

    • Stores anchor and grid geometry
    • Converts feet → pixels correctly
    • Draws rotated grid polygons
    • Generates serpentine numbering

  • export.py

    • Writes ESRI Shapefiles
    • Fields: area_name, name, area, gid
    • Writes UTF-8 .cpg
    • Optional WGS84 support

Notable Safeguards

  • Explicit warnings for geographic CRS
  • Axis-specific pixel size handling (non-square pixels)
  • Block-level TIFF validation before display
  • No silent assumptions about units or projections

Tool 2: Orthomosaic Viewer & Cropper

Animated GIF showing cropper workflow

What this tool is

A small, self-contained GeoTIFF viewer and cropper for very large orthomosaic rasters.

It is optimized for:

  • Fast pan/zoom
  • Inspecting specific regions
  • Exporting exact viewport crops at native resolution

What it does

  • Opens large orthomosaic GeoTIFFs
  • Builds .ovr overviews when missing
  • Renders the visible window to a Tkinter canvas

  • Supports:

    • Mouse drag panning
    • Mouse wheel zooming

  • Displays:

    • Pixel coordinates
    • Map coordinates

  • Exports the current view via gdal.Translate

    • Uses srcWin
    • Preserves CRS, geotransform, nodata
    • Always exports full-resolution data

Entry Points

The logic is intentionally duplicated with different defaults:

  • main.py

    • Primary viewer
    • Extra startup logging
    • Default: Orthomosaic.data.tif

  • hort.py

    • Same GUI logic
    • Cleaner startup
    • Default: Orthomosaic.data.tif

  • hort_cropper.py

    • Same GUI logic
    • Default: Orthomosaic.data_HFII.tif

Launchers

  • run_ortho_cropper_gui.sh
  • run_hort_cropper_gui.sh

Both:

  • Launch the GUI
  • Log stdout/stderr to desktop/home
  • Surface GDAL errors cleanly

Display & Export Details

  • Non-uint8 rasters are normalized for display

    • Percentile stretch (1–99%)

  • Export always uses:

    • Original data type
    • Native resolution
    • Original projection

Data & Environment

Dependencies

  • Python
  • tkinter
  • Pillow
  • numpy
  • GDAL / OGR / OSR (osgeo)

Environment

  • Conda environment noted as geotiff_env

Data Directories

  • ROUTINE_IN, TEST_IN — input GeoTIFFs
  • ROUTINE_OUT, TEST_OUT — exported shapefiles and crops

You May Also Enjoy

Shiny-Based ANOVA Tooling

3 minute read

This repository documents the development of a set of Shiny-based ANOVA analysis tools, along with report templates and exploratory scripts used to refine bo...