CornLeaf-MIA Net: a novel deep learning architecture for corn leaf disease detection in precision agriculture
Abstract
Early detection of corn leaf diseases is important for maintaining crop productivity and supporting precision agriculture. This study proposes CornLeaf-multi-scale intelligent architecture network (CornLeaf-MIA Net), a lightweight convolutional neural network (CNN) for multiclass classification of blight, common rust, gray leaf spot, and healthy corn leaves. A dataset of 4,013 labeled images was preprocessed through resizing, normalization, and training-stage augmentation. The proposed model was evaluated against MobileNetV2, VGG16, and ResNet50 under identical experimental conditions. CornLeaf-MIA Net achieved 98.38% testing accuracy, with weighted precision, recall, and F1-score of 98.50%, 98.38%, and 98.40%, respectively. The model contains 2,982,052 trainable parameters and achieved a training-validation accuracy gap of 0.71%. It also outperformed MobileNetV2 (94.40%), VGG16 (94.89%), and ResNet50 (95.39%) in testing accuracy. The trained model was implemented in a Flask-based web application using OpenCV and TensorFlow for automated multi-leaf disease diagnosis. The results indicate that the proposed task-specific multi-scale CNN provides high classification performance with a relatively compact architecture and demonstrates practical applicability through web-based implementation.
Keywords
Computer vision; Convolutional neural network; Corn leaf disease detection; CornLeaf-MIA Net; Deep learning
Full Text:
PDFDOI: https://doi.org/10.11591/eei.v15i6.14038
Refbacks
- There are currently no refbacks.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Bulletin of Electrical Engineering and Informatics (BEEI)
ISSN: 2089-3191
,
e-ISSN: 2302-9285
This journal is published by the
Institute of Advanced Engineering and Science (IAES)
.