The invention describes a system for agricultural vehicles to automatically steer themselves along crop rows. It uses cameras to identify crop rows, specifically those that are raised planting beds, and then calculates the vehicle's angle and side-to-side position relative to the lane. The vehicle then uses this information to stay centered and follow the row, with one specific version using two crop rows and fitting lines to individual plants.
Why it matters: Filed before the widespread availability of highly efficient deep learning models for real-time image analysis on edge devices. The specific tasks of detecting crop rows, bounding boxes, and fitting lines to plants have become significantly more robust and accurate with advancements in computer vision since 2020, making the core software easier to implement reliably.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI