Project ECHO
Student research on a crop-health survey drone, where I lead the software, with a Python and OpenCV prototype that estimates plant stand coverage.
- role
- Software Lead
- team
- Three-person agricultural UAV research team, Purdue University
- year
- 2026
- stack
- Python, OpenCV
Problem
Project ECHO is a student research drone that photographs farm fields and analyzes the images for crop health. I lead the software on a team of three, alongside an aerospace engineering student and an electrical engineering student. We wanted a project that combined hardware and software, and Purdue’s agriculture department began supporting it in fall 2026.
What I did
- Secured $5,000 in research funding with our faculty advisors.
- Built a Python and OpenCV prototype that estimates plant stand coverage from pixel color, and tested it on thousands of crop images from private and public datasets.
- Designed the software pipeline: GPS mapping of the field, an automated flight path that captures images at set intervals, crop-health analysis, and a 3D reconstruction of the field from the overlapping photos.
- Chose a standard RGB camera over a multispectral one. It costs far less and still gives accurate results, and multispectral is an option for a later version.
Result
The project is ongoing. The main challenge is integration: the flight controller, mapping software and analysis tools come from different sources, some of them dated, and they all have to work together within a tight budget and weight limit without losing data accuracy.