English

中国农机化学报

中国农机化学报 ›› 2024, Vol. 45 ›› Issue (9): 215-219.DOI: 10.13733/j.jcam.issn.2095-5553.2024.09.033

• 农业信息化工程 • 上一篇    下一篇

基于 Mask R-CNN的复杂环境下辣椒识别方法研究

付晓鸽 1,2,李涵 1,2,左治江 1,2,杜铮 3   

  1. (1.江汉大学精细爆破国家重点实验室,武汉市,430056;2.爆破工程湖北省重点实验室,武汉市,430056;3.武汉市农业科学院农业机械化研究所,武汉市,430207)
  • 出版日期:2024-09-15 发布日期:2024-09-04
  • 基金资助:
    湖北省教育厅百校联百县 —高校服务乡村振兴科技支撑行动计划(BXLBX0369);武汉市知识创新专项曙光计划项目(2022010801020378)

Research on pepper recognition method in complex environment based on Mask R-CNN 

Fu Xiaoge1,2,Li Han1,2,Zuo Zhijiang1,2,Du Zheng3   

  1. (1. State Key Laboratory of Percision Blasting,Jianghan University,Wuhan,430056,China; 2. Hubei Key Laboratory of Blasting Engineering,Wuhan,430056,China; 3. Institute of Agricultural Mechanization,Wuhan Academy of Agricultural Sciences,Wuhan,430207,China) 
  • Online:2024-09-15 Published:2024-09-04

摘要:

摘要:针对辣椒采摘机器人在真实场景中辣椒簇状、粘连和光照不均导致无法精准采摘辣椒的问题,提出一种基于 Mask R-CNN实例分割网络模型的辣椒识别方法。以真实场景下的辣椒为研究对象,采集自然生长的辣椒图像 4 496张,对其中的 4 000张进行数据标注作为数据集,通过设置不同的学习率、训练周期和模型网络层对数据集进行训练。试验结果表明,Mask R-CNN网络模型对真实场景下辣椒的识别和分割效果较好,平均准确率达到 90. 34%,平均速度达到 0. 82 s/幅,为智能辣椒采摘机器人的辣椒分割识别和定位提供有力的技术支撑。

关键词: 辣椒识别, 实例分割, Mask R-CNN, 神经网络, 采摘机器人

Abstract:

In order to solve the problem that pepper picking robots can not pick pepper accurately in real scenes due to pepper clusters, adhesion and uneven lighting, a pepper recognition method based on Mask R-CNN instance segmentation network model is proposed. With pepper in the real scene as the research object,4 496 images of naturally growing pepper were collected,and 4 000 of them were labeled as data sets. The data sets were trained by setting different learning rates,training cycles and model network layers. The experimental results show that the Mask R-CNN network model has a good effect on pepper recognition and segmentation in the real scene,with an average accuracy of 90. 34% and an average speed of 0. 82 s/frame,providing a strong technical support for pepper segmentation recognition and location of intelligent pepper picking robot.

Key words: pepper recognition;instance segmentation;Mask R-CNN;neural networks;picking robot ,

中图分类号: