Grading Tomatoes with an ESP32 and ML

An ESP32 running machine learning can now automatically grade tomatoes, thanks to a system that recently appeared on Hackaday. The project loads different trained models based on whether it's processing standard or cherry tomatoes, then analyzes optical sensor data to evaluate color consistency and size for each fruit. The program includes heuristic validity checks—it knows the approximate dimensions of each variety and flags anything that seems too large for cherry tomatoes or too small for standard ones, reducing but not eliminating miscategorization. The code is memory-intensive and requires a larger stack, but you can experiment with a web-based simulator without needing hardware or actual produce.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton