欢迎您访问《智慧农业(中英文)》官方网站! English

Smart Agriculture ›› 2021, Vol. 3 ›› Issue (4): 99-110.doi: 10.12133/j.smartag.2021.3.4.202109-SA006

• 信息处理与决策 • 上一篇    下一篇

改进YOLOv4的温室环境下草莓生育期识别方法

龙洁花1,2(), 郭文忠1, 林森1(), 文朝武1, 张宇1, 赵春江1   

  1. 1.北京市农林科学院智能装备技术研究中心,北京 100097
    2.上海海洋大学 信息学院,上海 201306
  • 收稿日期:2021-09-14 修回日期:2021-11-08 出版日期:2021-12-30
  • 基金项目:
    北京市科技计划(Z211100004621006);北京市农林科学院青年基金(QNJJ202027);宁夏回族自治区重点研发计划项目(2018BBF02024)
  • 作者简介:龙洁花(1997-),女,硕士研究生,研究方向为农业机器人和机器视觉。E-mail:seven060422@163.com
  • 通信作者: 林 森(1984-),男,高级工程师,研究方向为农业智能装备。电话:15652795990。E-mail:linseng@nercita.org.cn

Strawberry Growth Period Recognition Method Under Greenhouse Environment Based on Improved YOLOv4

LONG Jiehua1,2(), GUO Wenzhong1, LIN Sen1(), WEN Chaowu1, ZHANG Yu1, ZHAO Chunjiang1   

  1. 1.Beijing Academy of Agriculture and Forestry Sciences Intelligent Equipment Technology Research Center, Beijing 100097, China
    2.College of Information Science, Shanghai Ocean University, Shanghai 201306, China
  • Received:2021-09-14 Revised:2021-11-08 Online:2021-12-30
  • Foundation items:Beijing Municipal Science and Technology Plan (Z211100004621006); Beijing Academy of Agricultural and Forestry Sciences Youth Foundation (QNJJ202027); Key Research and Development Plan Project of the Ningxia Hui Autonomous Region (2018BBF02024)
  • About author:LONG Jiehua, E-mail:seven060422@163.com
  • Corresponding author:LIN Sen, E-mail:linseng@nercita.org.cn

摘要:

针对目前设施农业数字化栽培调控技术中对作物的生育期实时检测与分类问题,提出一种改进YOLOv4的温室环境下草莓生育期识别方法。该方法将注意力机制引入到YOLOv4主干网络的跨阶段局部残差模块(Cross Stage Partial Residual,CSPRes)中,融合草莓不同生长时期的目标特征信息,同时降低复杂背景的干扰,提高模型检测精度的同时保证实时检测效率。以云南地区的智能设施草莓为试验对象,结果表明,本研究提出的YOLOv4-CBAM(YOLOv4-Convolutional Block Attention Module)模型对开花期、果实膨大期、绿果期和成熟期草莓的检测平均精度(Average Precision,AP)分别为92.38%、82.45%、68.01%和92.31%,平均精度均值(Mean Average Precision,mAP)为83.79%,平均交并比(Mean Inetersection over Union,mIoU)为77.88%,检测单张图像时间为26.13 ms。YOLOv4-CBAM模型检测草莓生育期的mAP相比YOLOv4、YOLOv4-SE、YOLOv4-SC模型分别提高8.7%、4.82%和1.63%。该方法可对草莓各生育期目标进行精准识别和分类,并为设施草莓栽培的信息化、规模化调控提供有效的理论依据。

关键词: 目标检测, 草莓, 生育期识别, YOLOv4, 残差模块, 注意力机制, 损失函数

Abstract:

Aiming at the real-time detection and classification of the growth period of crops in the current digital cultivation and regulation technology of facility agriculture, an improved YOLOv4 method for identifying the growth period of strawberries in a greenhouse environment was proposed. The attention mechanism into the Cross Stage Partial Residual (CSPRes) module of the YOLOv4 backbone network was introduced, and the target feature information of different growth periods of strawberries while reducing the interference of complex backgrounds was integrated, the detection accuracy while ensured real-time detection efficiency was improved. Took the smart facility strawberry in Yunnan province as the test object, the results showed that the detection accuracy (AP) of the YOLOv4-CBAM model during flowering, fruit expansion, green and mature period were 92.38%, 82.45%, 68.01% and 92.31%, respectively, the mean average precision (mAP) was 83.78%, the mean inetersection over union (mIoU) was 77.88%, and the detection time for a single image was 26.13 ms. Compared with the YOLOv4-SC model, mAP and mIoU were increased by 1.62% and 2.73%, respectively. Compared with the YOLOv4-SE model, mAP and mIOU increased by 4.81% and 3.46%, respectively. Compared with the YOLOv4 model, mAP and mIOU increased by 8.69% and 5.53%, respectively. As the attention mechanism was added to the improved YOLOv4 model, the amount of parameters increased, but the detection time of improved YOLOv4 models only slightly increased. At the same time, the number of fruit expansion period recognized by YOLOv4 was less than that of YOLOv4-CBAM, YOLOv4-SC and YOLOv4-SE, because the color characteristics of fruit expansion period were similar to those of leaf background, which made YOLOv4 recognition susceptible to leaf background interference, and added attention mechanism could reduce background information interference. YOLOv4-CBAM had higher confidence and number of identifications in identifying strawberry growth stages than YOLOv4-SC, YOLOv4-SE and YOLOv4 models, indicated that YOLOv4-CBAM model can extract more comprehensive and rich features and focus more on identifying targets, thereby improved detection accuracy. YOLOv4-CBAM model can meet the demand for real-time detection of strawberry growth period status.

Key words: object detection, strawberry, growth period recognition, YOLOv4, residual module, attention mechanism, loss function

中图分类号: