YOLO-RDS: An efficient algorithm for monitoring the uprightness of seedling transplantation

文献类型： 外文期刊

第一作者： Jin, Xin

作者： Jin, Xin;Zhu, Xiaowu;Li, Mingyong;Li, Shaofan;Ji, Jiangtao;Jin, Xin;Xiao, Liqiang;Zhao, Bo;Ji, Jiangtao

作者机构：

关键词： Mechanized transplanting; Seedlings; Target detection; Uprightness identification; Monitoring system

期刊名称：COMPUTERS AND ELECTRONICS IN AGRICULTURE （ 影响因子：8.3； 五年影响因子：8.3 ）

ISSN： 0168-1699

年卷期： 2024 年 218 卷

页码：

收录情况： SCI

摘要： During the mechanized transplanting process, the planting angle and standing posture of seedlings are essential factors in evaluating the survival of the planting, and the uprightness of the planting is an essential indicator in evaluating the planting quality. With the gradual expansion of mechanization, the traditional method of manually checking the uprightness of mechanized planting has gradually become inapplicable. To address this problem, an online monitoring system for the uprightness of transplanting machines based on machine vision and deep learning is constructed. The system mainly consists of an image acquisition device, a vehicle -mounted computer, and an interactive interface. Furthermore, to improve the identification errors caused by the unstructured environment in the field, a field planting uprightness monitoring device was designed to integrate LED light sources, photoelectric sensors, a CCD camera, and light hoods. An efficient algorithm YOLO-RDS for the detection of the planting posture of seedlings is proposed using a single -stage detection paradigm. First, the traditional (HBB) target detection box is improved and the circle smooth label is introduced to adapt to different seedling planting inclined postures. Secondly, the RWBF detection head network was constructed, and the stem incline angle was extracted using adaptive rotation to calculate the actual incline angle of the seedlings in threedimensional space. Experimental results show that YOLO-RDS is better than the classic rotating target detection method, the average accuracy (mAP) is 97%, and FPS is 22. By encapsulating the algorithm, a seedling planting uprightness monitoring system was integrated and developed. Field experiments show that the average relative error rate of uprightness identification monitored by the system is 3.09%, the accuracy rate is 92.56%, and the absolute angle error is 0 degrees -3 degrees. The system has good accuracy and stability, meets the requirements for seedling uprightness monitoring, and can provide technical support for the quality evaluation of seedling transplanter operations.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序