Ginger direction adjusting and plant spacing control method and system based on instance segmentation

By using an instance-based segmentation method and a depth map repair algorithm, the problem of low efficiency in manual orientation during ginger planting was solved, achieving accurate identification and automatic orientation of ginger, thus improving planting efficiency and quality.

CN121330241AActive Publication Date: 2026-01-13BEIJING UNIV OF POSTS & TELECOMM
6 Cites 2 Cited by

Patent Information

Application Number
CN202511597088.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-04
Publication Date
2026-01-13
Estimated Expiration
2045-11-04

AI Technical Summary

Technical Problem

Existing technologies rely on manual visual judgment and manual rotation during ginger planting, which is labor-intensive, inefficient, and inconsistent. Furthermore, the lack of effective instance segmentation methods makes it difficult to achieve automatic orientation and plant spacing control for ginger.

Method used

An instance-based segmentation method was adopted. A ginger seed dataset was collected using a depth camera, and an improved YOLO11-seg model was trained to obtain the mask information of ginger pieces and ginger shoots. The three-dimensional coordinates were calculated by combining the depth map repair algorithm, and a multi-factor scoring function was used to prioritize the seed picking operation and determine the adjustment angle and plant spacing correction.

Benefits of technology

It enables precise identification and automatic orientation of ginger, reduces labor costs, improves sowing efficiency and quality, and meets the needs of precision agriculture.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention belongs to the technical field of agricultural image recognition and intelligent direction adjustment, and discloses a ginger direction adjustment and plant spacing control method and system based on instance segmentation, and the method comprises the steps: training improved YOLO11-seg through collecting a stacked ginger seed data set, and obtaining the mask information of ginger blocks and ginger buds; and obtaining the three-dimensional coordinates of each target ginger seed in a camera coordinate system in combination with the depth image after the repair algorithm is used. According to the method, seed taking priority ranking is carried out on a plurality of ginger targets based on a multi-factor scoring function, and the direction adjusting angle and the plant spacing correction distance for controlling ginger seed placement are solved through the relative position relation between ginger buds and ginger blocks. According to the method, the ginger stacked in the box can be accurately identified, the direction adjusting placement parameters are obtained, and the method can be used in agricultural scenes such as intelligent seeding, so that manual seeding is replaced, the labor cost is reduced, the seeding efficiency and quality are improved, and the fine agriculture requirements are met.
Need to check novelty before this filing date? Find Prior Art