In just a few weeks, the company gained a reliable solution to digitize their crop monitoring. What was once a manual and error-prone process became a real-time, automated system, saving labor time and improving decision-making with accurate, structured data.
🇬🇧
AI-powered harvest monitoring
G’s Fresh, a UK-based agricultural company, sought to modernize how they measure and classify harvests like pumpkins and onions. Their goal was to leverage AI-powered computer vision to automatically detect, count, and categorize crops during harvesting.
Challenge
Traditionally, measuring crop yield by size and quantity required manual labor and estimations. G’s Fresh wanted to automate the classification of pumpkins and onions by size and type, generate accurate real-time data during harvest, and export structured reports (CSV) for internal analysis. The complexity increased when moving from pumpkins, which are larger and easier to detect, to onions, which are smaller, more numerous, and harder to classify.
Solution
We developed a Python-based computer vision system integrated with a camera mounted on harvesting equipment. The system detects pumpkins and onions in real time, classifies them by size, generates live counts during harvesting, and outputs a CSV with total quantities and classifications.
Outcome
