/WheatLeafDiseaseByPyTorch

PyTorch CNN for classify Wheat Leaf Disease

Primary LanguageJupyter Notebook

CNN based classification of Wheat Leaf Disease

GitHub issues Github Stars visitor badge


🔹Introduction

Using prevalent CNN to classify 4 class of Wheat Leaf Disease(Brown rust,Healthy,Septoria,Yello rust).

🔹DataSet

https://drive.google.com/file/d/1dtoVdpNf-v6__dmSy3lUpLYKoCOhEYUw/view?usp=drive_link

🔹Environment

For free product by Google Colab

🔹Refer

https://www.mdpi.com/2072-4292/14/14/3446

🔹Result

Memory Used

AlexNet 217.5149078 MB
VGG11 491.2881012 MB
GoogleNet 21.37757874 MB
ResNet18 42.64283752 MB
DenseNet121 26.54249573 MB
EfficientNet_b7 243.3670197 MB
MobilenetV3Large 16.23838806 MB

Time use

AlexNet 109.667897 second
VGG11 161.1042612 second
GoogleNet 140.4423647 second
ResNet18 122.6961145 second
DenseNet121 311.7152424 second
EfficientNet_b7 537.789294 second
MobilenetV3Large 132.0254643 second

ACC

AlexNet 95.1923077%
VGG11 96.6346154%
GoogleNet 92.7884615%
ResNet18 94.7115385%
DenseNet121 96.3942308%
EfficientNet_b7 92.3076923%
MobilenetV3Large 96.0336538%

All the above results were obtained with the pre-trained model

🔹Run

1.cropImg.py
2.wheatLeafDisease_delineation.py
3.wheatLeafDiesas.ipynb

Note:Keep datasets and programs in the same directory