OrthoFly

This repository contains OrthoFly, an initiative to develop advanced aerial mapping software for the AirCRAFT Lab’s Student Unmanned Aerial Systems (SUAS) team at Saint Louis University. The team participates annually in the SUAS competition hosted by RoboNation.

OrthoFly is designed to autonomously capture, generate, and export high-resolution orthomosaic maps of mission areas using geotagged aerial imagery.

The current working version relies on OpenDroneMap for post-processed image stitching and map generation. In parallel, custom software is under active development to enable in-flight orthomosaic creation and near real-time mapping capabilities.

Mission:

The mission for this portion of the competition is to design and implement a system that enables a UAS to map a 10-acre area with high accuracy, generating a complete orthomosaic map and exporting it to a USB flash-drive within a 30-minute mission window.

For further information about missions and tasks, please refer to the 2025 SUAS competition handbook.

Competition Information: SUAS (Student Unmanned Aerial Systems) 2025

The SUAS competition is an annual event that encourages students to design, integrate, and demonstrate a UAS capable of autonomous flight and navigation. The competition engages students in real-world missions that simulate UAS tasks, including remote sensing and autonomous decision-making. This year it is held from June 24-26, 2025 at St. Mary’s County Regional Airport, California, MD.

Mission Task:

• Mapping Task: Teams must generate high-resolution imagery maps of a 10-acre area, which will be evaluated for coverage, accuracy, and stitching quality. The mapping must be completed within the 30-minute mission window, and maps are submitted via USB flash-drive.

Mapping Scoring Criteria:

• Coverage: The completeness of the mapped area.
• Projection Accuracy: How well the map matches the real-world coordinates.
• Stitching Quality: The quality of stitching between images (without noticeable seams or misalignments).
• Overall Quality: The map's usability as a professional-grade map.

Project Components

1. Capture:

• Use of on-board camera to capture images of the mapping region. Images are then ran trhough geotagging software (using GPS and IMU data).

2. Process:

• The images are processed and stitched using mapping software OpenDroneMap through the python wrapper PyODM.

3. Export:

• The final map will be exported to a USB flash drive for submission during the mission.

Tech Stack

• OpenDroneMap (ODM): A popular open-source tool used for photogrammetry and creating 3D models from images. The system will leverage PyODM (a Python wrapper) for automation.
• Python: Primary programming language for integrating the components and implementing the mapping logic.

Hardware

• UAS (Plane): DragonFly UAV, equipped with a flight controller and camera for autonomous operation.
• On-board Computer: NVIDIA Jetson Orin Nano for processing captured images and running the required algorithms.
• Flight Controller: Pixhawk 4 to manage the UAV’s flight.
• On-boad Camera: Captures images for processing.
• Off-Board Server: Ground PC to manage and monitor the UAV’s flight and data processing.

Demo Instructions

This project currently supports Windows OS only (Linux support coming soon).

Follow these steps to install dependencies and test-run the software with provided sample geotagged images.

Prerequisites

• Python 3 (latest version recommended)
• Docker Desktop installed and running (Docker installation guide)
• NodeODM API running (see NodeODM setup guide)

Setup

1. Clone the repository

   git clone https://github.com/yourusername/SUAS_Mapping.git
   cd SUAS_Mapping

2. Create and activate a virtual environment

   python -m venv venv
   venv\Scripts\activate

3. Install Python dependencies

   pip install -r requirements.txt

4. Start the NodeODM API

   Open a separate terminal window and launch NodeODM. (Refer to the NodeODM setup guide for detailed instructions.)

Running the Demo with Sample Images

1. Prepare the sample data

   Copy the provided sample images from:

   SUAS_Mapping/data/sample_data/fourth_ave_test_data
   

   into:

   SUAS_Mapping/data/raw
   

2. Run the mapping script

   python odm_mapping.py

3. View the output

   The processed map outputs will be saved to:

   SUAS_Mapping/data/output
   

Using Your Own Geotagged Images

To test with your own images:

• Place them in the data/raw folder.
• Follow the same steps to run the script.

Troubleshooting NodeODM

If you encounter issues with NodeODM or PyODM please refer to the their respective guides:

• NodeODM Guide
• PyODM Documentation

In Development

• SITL test and setup using Gazebo simulation (work in progress).

Reports for Analysis

All generated outputs in the data/output folder will include a task report that can be reviewed or exported for any analysis use.

AirCRAFT Lab at Saint Louis University

This project is being developed under the Aircraft Computational & Resource Aware Fault Tolerance (AirCRAFT) Lab at Saint Louis University. The AirCRAFT Lab's mission is to address aircraft safety through cutting-edge, high-risk/high-reward flight control algorithms, particularly in adverse conditions such as sensor and actuator failures.

Contact

For inquiries regarding this research, please reach out to:

• Aida Bah (SUAS Computer Science Division Lead) – [email protected]

• Dr. Srikanth Gururajan (Lab Director) – [email protected]

License

This project is licensed under the AirCRAFT Lab Research License (ALRL) v1.0. It is for internal research use at the AirCRAFT Lab, Saint Louis University (SLU). External distribution, commercial use, and public disclosure are prohibited without prior approval.

For full details, see the file LICENSE.md.