基于改进SlowFast模型的设施黄瓜农事行为识别方法  

Recognition Method of Facility Cucumber Farming Behaviours Based on Improved SlowFast Model

作　　者：何峰[1,2] 吴华瑞 史扬明 朱华吉[1,2,3,4] HE Feng;WU Huarui;SHI Yangming;ZHU Huaji(School of Computer Science and Communication Engineering,Jiangsu University,Zhenjiang 212013,China;National Engineering Technology Research Centre for Agricultural Informatisation,Beijing 100097,China;Information Technology Research Centre,Beijing Academy of Agriculture and Forestry,Beijing 100097,China;Key Laboratory of Digital Rural Technology,Ministry of Agriculture and Rural Development,Beijing 100097,China)

机构地区：[1]江苏大学计算机科学与通信工程学院,江苏镇江212013 [2]国家农业信息化工程技术研究中心,北京100097 [3]北京市农林科学院信息技术研究中心,北京100097 [4]农业农村部数字乡村技术重点实验室,北京100097

出　　处：《智慧农业（中英文）》2024年第3期118-127,共10页Smart Agriculture

基　　金：中央引导地方科技发展资金项目(2023ZY1-CGZY-01);财政部和农业农村部:国家现代农业产业技术体系资助(CARS-23-D07)。

摘　　要：[目的/意义]农事行为活动识别对设施蔬菜生产精准化调控有着重要意义,在一定程度上可以通过查看农事操作的时间、操作过程是否合理来减少因农事行为不当导致产量下降。为了解决农事行为识别方法中由于黄瓜叶片和设施遮挡导致识别准确率不高的问题,提出一种名为SlowFast-SMC-ECA (SlowFast-Spatio-Temporal Excitation、Channel Excitation、Motion Excitation-Efficient Channel Attention)的农事活动行为识别算法。[方法]该算法主要基于SlowFast模型,通过改进Fast Pathway和Slow Pathway中的网络结构来提高对于农事活动中手部动作特征和关键特征的提取能力。在Fast Pathway中,引入多路径激励残差网络的概念,通过在信道之间插入卷积操作来增强它们在时域上的相互关联性,从而更好地捕捉快速运动信息的细微时间变化。在Slow Pathway中,将传统的Residual Block替换为ECA-Res结构,以提高对通道信息的捕获能力。这两项改进有效地加强了通道之间的联系,提升了特征之间的语义信息传递,进而显著提升了农事行为识别的准确率。此外,为了解决数据集中类别不均衡的问题,设计了平衡损失函数(Smoothing Loss),通过引入正则化系数,平衡损失函数可以有效地处理数据集中的类别不均衡情况,提高模型在各个类别上的表现。[结果和讨论]改进的SlowFast-SMC-ECA模型在农事行为识别中表现出良好的性能,各类行为的平均识别精度达到80.47%,相较于原始的SlowFast模型有约3.5%的提升。[结论]本研究在农事行为识别中展现出良好的性能。这对农业生产的智能化管理和决策具有重要意义。[Objective]The identification of agricultural activities plays a crucial role for greenhouse vegetables production,particularly in the precise management of cucumber cultivation.By monitoring and analyzing the timing and procedures of agricultural operations,effective guidance can be provided for agricultural production,leading to increased crop yield and quality.However,in practical applications,the recognition of agricultural activities in cucumber cultivation faces significant challenges.The complex and ever-changing growing environment of cucumbers,including dense foliage and internal facility structures that may obstruct visibility,poses difficulties in recognizing agricultural activities.Additionally,agricultural tasks involve various stages such as planting,irrigation,fertilization,and pruning,each with specific operational intricacies and skill requirements.This requires the recognition system to accurately capture the characteristics of various complex movements to ensure the accuracy and reliability of the entire recognition process.To address the complex challenges,an innovative algorithm:SlowFast-SMC-ECA(SlowFast-Spatio-Temporal Excitation,Channel Excitation,Motion Excitation-Efficient Channel Attention)was proposed for the recognition of agricultural activity behaviors in cucumber cultivation within facilities.[Methods]This algorithm represents a significant enhancement to the traditional SlowFast model,with the goal of more accurately capturing hand motion features and crucial dynamic information in agricultural activities.The fundamental concept of the SlowFast model involved processing video streams through two distinct pathways:the Slow Pathway concentrated on capturing spatial detail information,while the Fast Pathway emphasized capturing temporal changes in rapid movements.To further improve information exchange between the Slow and Fast pathways,lateral connections were incorporated at each stage.Building upon this foundation,the study introduced innovative enhancements to both pathways,improv

关 键 词：农事活动行为 SlowFast模型 多路径激励残差网络 ECA-Res 平衡损失函数 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]