Automated Malabar Chestnut Planting Machine with Deep Learning Visual Recognition and Innovative Mechanisms

Cultivated for its ornamental appeal, the Malabar chestnut demands precise planting for optimal growth, emphasizing the necessity of down ward-facing seed germination points. Amidst a scarcity of agricultural labor, there is a growing demand for automated planting solutions. This paper presents an automatic planting machine for Malabar chestnut, utilizing deep learning image recognition to ensure proper seed orientation during planting. The machine incorporates a novel mechanism, leveraging high-speed pneumatic action and mechanical principles to guarantee accurate seed orientation. We provide insights into the architecture and training of the convolutional neural network-based recognizer, the design and analysis of the planting machine, and the system’s performance in field tests. Results from field tests affirm an $85 \%$ success rate in proper seed planting, achieving an average planting speed of one seed every 3 seconds.