首页|基于改进卷积神经网络的辣椒病虫害检测

基于改进卷积神经网络的辣椒病虫害检测

扫码查看
针对使用卷积神经网络对辣椒病虫害进行检测有参数多、计算量大和推理时间过长等问题,提出一种基于MobileNet-V2 改进的轻量化神经网络,将MobileNet-V2 的BN层中的激活函数全部替换为Leaky ReLU,保留特征图中更多的有效正负信息,以提高性能和减少计算复杂度,增强模型的鲁棒性.在公开的辣椒病虫害数据集上使用 VGG16、ResNet34 和MobileNet-V2等模型对比后,改进的 MobileNet-V2 表现出更高的准确性和更少的参数量.相对于原来的 MobileNet-V2 准确率提升 4%,相对 VGG16、ResNet34 两种模型参数分别下降 97%和 87%.能够移动端设备实现实时病虫害检测,提供高效便捷解决方案.
Pepper pest detection based on neural network
There are many problems in using convolutional neural network to detect pepper diseases and insect pests,such as large number of parameters,large amount of calculation and too long inference time.This paper proposes an improved lightweight neural network based on MobileNet-V2,replacing all activation functions in the BN layer of MobileNet-V2 with Leaky ReLU,retaining more effective positive and negative information in the feature map to improve performance and reduce Computational complexity and enhanced model robustness.After comparing models such as VGG16,ResNet34 and MobileNet-V2 on the public pepper diseases and insect pests data set,the improved MobileNet-V2 showed higher accuracy and fewer parameters.Compared with the original MobileNet-V2,the accuracy increased by 4%,and compared with VGG16 and ResNet34,the parameters of the two models dropped by 97%and 87%respectively.It can realize real-time pest and disease detection on mobile devices and provide efficient and convenient solutions.

pepper pests and diseasesVGG16MobileNet-V2ResNet34Leaky ReLU

史明健、袁缘、刘铭

展开 >

长春工业大学 数学与统计学院,吉林 长春 130012

辣椒病虫害 VGG16 MobileNet-V2 ResNet34 Leaky ReLU

吉林省发改委省预算内基本建设资金吉林省科技厅自然科学基金项目

2022C043-220200201157JC

2024

长春工业大学学报
长春工业大学

长春工业大学学报

影响因子:0.282
ISSN:1674-1374
年,卷(期):2024.45(3)