您好，欢迎访问北京市农林科学院 机构知识库！
北京市农林科学院机构知识库

全部

  • 全部
  • 标题
  • 学者
  • 机构
  • 关键词

Development and testing of an online detection system for impurity content in wheat for combine harvesters

文献类型： 外文期刊

作者： Zhao, Junjie 1 ; Yin, Yanxin 1 ; Shao, Mingxi 3 ; Wang, Qian 1 ; Wang, Feng 1 ; Lu, Longxin 1 ;

作者机构： 1.Beijing Acad Agr & Forestry Sci, Res Ctr Intelligent Equipment, Beijing 100097, Peoples R China

2.State Key Lab Intelligent Agr Power Equipment, Beijing 100097, Peoples R China

3.Qinghai Univ, Sch Mech Engn, Xining 810003, Peoples R China

关键词： Wheat; Impurity content; GAW-YOLO model; Online detection system; Operation parameters

期刊名称：MEASUREMENT （ 影响因子：5.6； 五年影响因子：5.4 ）

ISSN： 0263-2241

年卷期： 2025 年 256 卷

页码：

收录情况： SCI

摘要： Grain impurity content is an important indicator for optimizing combine harvester operation parameters. Affected by factors such as complex environments, light intensity, and harvester body vibration, existing wheat impurity content online detection devices and methods face issues like low detection accuracy and poor stability, making them unsuitable for field operations. To provide drivers with accurate and stable grain impurity content information, this paper proposes an online intelligent detection method for wheat impurity content and develops a detection system. First, a reliable wheat material acquisition device was developed to ensure effective online image sample collection. Second, a GAW-YOLO model for wheat impurity detection was designed based on the YOLOv8 network, incorporating Ghost convolutions, an Adaptive Feature Fusion (APFF) module, and Wise-IoU loss function. Traditional convolutions were replaced with Ghost convolutions to reduce model FLOPs and parameters, improving real-time detection. The APFF module enhances the model's ability to capture multi-scale wheat impurity information, enabling better focus on small or obscured impurities. The improved GAW-YOLO model achieved 97.2% accuracy in impurity recognition and a detection speed of 42 FPS on Jetson Orin NX. Finally, an online detection system for wheat impurity information in combine harvesters was developed and installed in the GM100 combine harvester for field testing, with a mean relative error of 9.13% between detected and manually counted impurity content. The real-time impurity content detection results can provide a basis for the driver or automatic control system to adjust operation parameters in time, thus improving the harvester's operation quality and efficiency.

  • 相关文献
作者其他论文 更多>>
置顶
购物车
反馈

© 京ICP备09089781号-29 主办单位：北京市农林科学院

学术资源： 中国农业科学院 农业专业知识服务系统 农科联盟资源共建共享平台 中国农业科技文献与信息服务平台 中国农业期刊集成服务平台