Satellite Image Downloader

Download high-resolution satellite/map imagery from Google Maps for any bounding box and save it as a fully georeferenced GeoTIFF (or PNG/JPEG).

Give it two corner coordinates (top-left and bottom-right), a zoom level, and an output path — the tool fetches all the required map tiles in parallel, stitches them into a single image, and writes out a GeoTIFF with correct WGS-84 georeferencing (or a plain image if you use a .png/.jpg extension).

• Satellite Image Downloader
  • Features
  • Requirements
  • Installation
  • Usage
    • Options
    • Examples
  • Map styles
  • Notes
  • License

Features

• Download any bounding box at zoom levels 0–22
• Concurrent tile downloads (configurable worker count)
• Output as GeoTIFF with correct geotransform + WGS-84 CRS, ready to open in QGIS/ArcGIS/GDAL
• Multiple map styles: satellite, satellite with labels, terrain, roadmap, labels only
• Supports both the standard Google Maps tile server and the China-specific server (GCJ-02 -> WGS-84 corrected)

Requirements

• Python 3.12+
• GDAL native library and development headers (required to build the gdal Python bindings)
• uv for dependency management

On Debian/Ubuntu, install the GDAL system packages first:

sudo apt-get install -y gdal-bin libgdal-dev python3-gdal

Make sure the installed GDAL version matches (or is compatible with) the version pinned in pyproject.toml.

Installation

Clone the repo first:

git clone <your-repo-url>
cd satellite-image-downloader

Then set up your environment using either uv (recommended) or plain python/pip.

Option A: using uv

./setup_env.sh

setup_env.sh creates a .venv with uv, activates it, and builds the project. Alternatively, run the steps manually:

uv venv --python=3.12
source .venv/bin/activate
uv sync

Option B: using plain python + pip

python3 -m venv .venv
source .venv/bin/activate   # On Windows: .venv\Scripts\activate
pip install -r requirements.txt

Usage

With uv:

uv run main.py \
  --top-left <lat,lon> \
  --bottom-right <lat,lon> \
  --zoom <0-22> \
  --output <path/to/output.tif>

With plain python (after activating your virtual environment):

python main.py \
  --top-left <lat,lon> \
  --bottom-right <lat,lon> \
  --zoom <0-22> \
  --output <path/to/output.tif>

Options

Flag	Required	Default	Description
--top-left	yes	-	Top-left corner as latitude,longitude
--bottom-right	yes	-	Bottom-right corner as latitude,longitude
--output	yes	-	Output file path (.tif for GeoTIFF, or .png/.jpg for plain image)
--zoom	no	22	Zoom level (0-22). Higher = more detail, more tiles, longer download
--style	no	s	Map style, see Map styles
--server	no	Google	Tile server: Google or Google China
--workers	no	16	Number of concurrent download workers

Examples

Download a satellite GeoTIFF over a small area in Singapore at zoom 18:

# uv
uv run main.py \
  --top-left 1.288221,103.673125 \
  --bottom-right 1.266984,103.699828 \
  --zoom 18 \
  --output ~/sat.tif

# plain python
python main.py \
  --top-left 1.288221,103.673125 \
  --bottom-right 1.266984,103.699828 \
  --zoom 18 \
  --output ~/sat.tif

Download roadmap-style imagery as a PNG:

# uv
uv run main.py \
  --top-left 1.288221,103.673125 \
  --bottom-right 1.266984,103.699828 \
  --zoom 16 \
  --style m \
  --output ~/map.png

# plain python
python main.py \
  --top-left 1.288221,103.673125 \
  --bottom-right 1.266984,103.699828 \
  --zoom 16 \
  --style m \
  --output ~/map.png

Map styles

Style	Description
m	Standard map
s	Satellite imagery
y	Satellite imagery with labels
t	Terrain
p	Terrain with labels
h	Labels only (transparent)

Notes

• Higher zoom levels and larger bounding boxes download significantly more tiles — expect longer download times and larger output files.
• GeoTIFF output is written using WGS-84 (EPSG:4326) coordinates, matching the corners you provide.
• This tool talks directly to Google's map tile endpoints; usage is subject to Google's terms of service.

License

No license specified yet. Add one (e.g. MIT) if you plan to open-source this repository.