Design of Sugarcane Orientation System Based on Machine Vision

This subject is based on the pre-cut sugarcane planting machine, combined with the agronomic requirements of the sugarcane buds to be planted in the horizontal direction during the sugarcane planting process, to study the sugarcane directional planting system. The cut cane seed segments are placed in the sugarcane planting machine seed box. The initial orientation of the cane buds on the cane seeds is different. Therefore, the combination of machine vision and mechanical structure is used to identify and adjust the direction of the cane buds on the cane seeds. An important step in the planting plan. Adjusting the centroid of the cane bud to the horizontal direction is the key point of the machine vision recognition scheme. Therefore, it is necessary to formulate a reasonable visual recognition scheme and evaluation index. By comparing different cane bud visual detection algorithms, it is found that the convolutional neural network is more suitable for the cane bud. The recognition and positioning of the sugarcane has a success rate of 90%, which provides a research foundation for the intelligent and precise planting of sugarcane.