WALDO

WALDO — Whereabouts Ascertainment for Low-lying Detectable Objects — is an open-source AI system for finding small objects in drone and aerial imagery. It processes high-resolution aerial photos by tiling them and running detection models over each tile, so tiny targets like vehicles, people or debris that would vanish in a downscaled full-frame pass are still found and mapped to coordinates. The project has around 1,700 stars and was last updated in 2025, and its author describes it as state-of-the-art among free and open-source tools for drone AI.

Overview

Standard object detectors fail on drone footage because the interesting objects occupy a handful of pixels. WALDO’s pipeline is built around that reality: large images are sliced into overlapping tiles, detections are merged back into the full-frame coordinate space, and results can be reviewed and exported for mapping workflows such as search support, surveying and wildlife or infrastructure monitoring.

Key Features

• Tiling pipeline that detects small objects in high-resolution aerial images
• Detections mapped back to full-image coordinates for GIS-style use
• Tuned for drone imagery: vehicles, people and other low-lying targets
• Configurable models and tile sizes for different altitudes and cameras
• Runs locally on your own hardware — imagery never leaves your machine
• Free and open source, aimed at practical field use

💳 Pricing

Free. WALDO is released as free open-source software; you download and run it locally at no cost.

Who Is It For

Drone operators, search-and-rescue volunteers, surveyors and researchers who need to locate small objects in aerial imagery without paying for enterprise photogrammetry AI platforms.