Key Features

Essential features and capabilities of SightLinks

Input & Detection

Input Formats

Supports .jpg/.jgw files and .tif files with automatic zip extraction

Detection Models

Advanced YOLO-based models with configurable thresholds

Processing & Output

Georeferencing

Converts pixel coordinates to geographical coordinates

Output Options

Multiple output formats (JSON/TXT) with visualization options

System Requirements: Python 3.8+ required

Technical Details

Understanding the input formats and system configuration

Input Formats

Image Data (input_type=0)

  • Supported: ZIP, JPG, JPEG, PNG, JGW
  • ZIP files with JPG/PNG/JPEG + JGW
  • Individual JPG/PNG/JPEG + JGW files
  • Automatic ZIP extraction
  • Hidden files ("._") ignored

GeoTIFF Data (input_type=1)

  • Supported: ZIP, TIF, TIFF
  • ZIP files with GeoTIFF files
  • Individual GeoTIFF files
  • Automatic ZIP extraction
  • Hidden files ("._") ignored

Configuration

File Management

  • Each task creates a unique session ID: timestamp_uuid
  • Files are uploaded to: input/{session_id}/
  • Processing results saved to: run/output/{session_id}/
  • Results are compressed into a ZIP file after processing
  • Original directories are deleted after successful compression
  • ZIP file integrity is verified after creation

Parameters

  • input_type: '0' (Image) or '1' (GeoTIFF)
  • classification_threshold: Float, default '0.35'
  • prediction_threshold: Float, default '0.5'
  • save_labeled_image: 'true'/'false'
  • output_type: '0' (JSON) or '1' (TXT)
  • yolo_model_type: 'n', 's', or 'm'

Setup Steps

Follow these steps to get SightLinks up and running on your system

1. Clone Repository

Clone the SightLinks repository to your local machine.

git clone https://github.com/UCL-SightLinks/SightLinks-Main.git && cd SightLinks-Main
Make sure you have Git installed and configured

2. Virtual Environment

Set up a Python virtual environment for isolated dependencies.

python -m venv venv
Activate Environment
# On macOS/Linux:
source venv/bin/activate

# On Windows:
venv\Scripts\activate

3. Install Dependencies

Install all required Python packages for SightLinks.

# On Windows and Linux machines
sudo apt install update
sudo apt install gdal-bin libgdal-dev
pip install -r requirements.txt
# On MacOS machines
brew update
brew install gdal
pip install -r requirements.txt
This may take several minutes. Requires Python 3.8 or higher.

4. Run SightLinks

Place your input files and run the system.

python run.py
Place input files in the input/ directory before running
Example Configuration
{
  "uploadDir": "input",
  "inputType": "0",
  "classificationThreshold": 0.35,
  "predictionThreshold": 0.5,
  "saveLabeledImage": false,
  "outputType": "0",
  "yolo_model_type": "n"
}