基于深度网络集成的复杂背景甘蔗叶片病害识别  

Recognition of Sugarcane Leaf Diseases in Complex Backgrounds Based on Deep Network Ensembles

作　　者：马巍巍[1] 陈悦 王咏梅 MA Weiwei;CHEN Yue;WANG Yongmei(School of Computer Science and Artificial Intelligence,Hefei Normal University,Hefei 230061,China)

机构地区：[1]合肥师范学院计算机与人工智能学院,安徽合肥230061

出　　处：《智慧农业(中英文)》2025年第1期136-145,共10页Smart Agriculture

基　　金：国家自然科学基金(62071001);合肥师范学院大学生创新创业训练计划资助项目(S202414098181);合肥师范学院校级科研项目(2024KYJX44);安徽省社会科学创新发展研究课题(2023KY016)。

摘　　要：[目的/意义]农作物病害图像的随机性和复杂性仍给病害识别带来诸多挑战。针对自然条件下甘蔗叶片病害识别难题,本研究提出XEffDa模型。[方法]该模型利用色调、饱和度、亮度(Hue-Saturation-Value,HSV)颜色空间的图像分割与边缘处理技术去除背景干扰,根据特征融合策略,集成高效网络B0版本(Efficient Network B0,EfficientNetB0)、深度可分离卷积网络(Extreme Inception,Xception)和密集连接卷积网络201(Dense Convolutional Network 201,DenseNet201)作为特征提取器,采用预训练权重,通过贝叶斯优化确定顶层超参数,改进弹性网络(ElasticNet)正则化方法并加入随机失活(Dropout)层,以双重机制遏制过拟合现象。在甘蔗叶片病害数据集上训练并完成分类任务。[结果和讨论]模型集成后的识别准确率为97.62%,对比EfficientNetB0、Xception单模型及EfficientNetB0与其他深度网络结合模型识别准确率分别提高了9.96、6.04、8.09、4.19、1.78个百分点。融合实验进一步表明,加入改进ElasticNet正则化后的网络较主干网络其准确率、精确度、召回率及F1值分别提高了3.76、3.76、3.67及3.72个百分点。最大概率散点图结果显示预测最大概率值不低于0.5的比例高达99.4%。[结论]XEffDa模型具有更好的鲁棒性和泛化能力,能为农作物叶片病害精准防治提供参考。[Objective]Sugarcane is an important cash crop,and its health status affects crop yields.However,under natural environmental conditions,the identification of sugarcane leaf diseases is a challenging problem.There are various issues such as disease spots on sugarcane leaves being blocked and interference from lighting,which make it extremely difficult to comprehensively obtain disease information,thus significantly increasing the difficulty of disease identification.Early image recognition algorithms cannot accurately extract disease features and are prone to misjudgment and missed judgment in practical applications.To solve the problem of identifying sugarcane leaf diseases under natural conditions and break through the limitations of traditional methods,a novel identification model,XEffDa was proposed in this research.[Methods]The XEffDa model proposed implemented a series of improvement measures based on the ensemble learning framework,aiming to significantly improve the accuracy of classifying and identifying sugarcane leaf diseases.Firstly,the images in the sugarcane leaf disease dataset under natural conditions were pre-processed.Real-time data augmentation techniques were used to expand the scale of the dataset.Meanwhile,HSV image segmentation and edge-processing techniques were adopted to effectively remove redundant backgrounds and interference factors in the images.Considering that sugarcane leaf disease images were fine-grained images,in order to fully extract the semantic information of the images,the transfer learning strategy was employed.The pre-trained models of EfficientNetB0,Xception,and DenseNet201 were loaded respectively,and with the help of the pre-trained weight parameters based on the ImageNet dataset,the top layers of the models were frozen.The performance of the validation set was monitored through the Bayesian optimization method,and the parameters of the top-layer structure were replaced,thus achieving a good balance between op‐timizing the number of model parameters and the overall p

关 键 词：甘蔗叶片病害 图像识别 EfficientNet Xception DenseNet201 模型集成 

分 类 号：S24[农业科学—农业电气化与自动化]