Abstract:
The increasing number of potato leaf diseases offers a danger to global food security,
necessitating creative technologies for rapid and precise identification. This work proposes
a novel approach leveraging advanced deep learning algorithms with a particular focus on
the DenseNet model, which achieved the greatest accuracy of 87% in classifying various
potato diseases. This research intentionally picked nine common potato illnesses for this
study: Bacteria, Early Blight, Fungi, Healthy, Late Blight, Nematode, Pest, Phytophthora
and Virus. The DenseNet model's superior feature extraction capabilities enable more
precise and reliable illness diagnosis compared to older methods. By utilizing the
DenseNet architecture exploited its potential to capture complex patterns and hierarchical
characteristics in the picture data, considerably boosting the model’s diagnostic accuracy.
This innovation offers a powerful tool for farmers enabling early identification and
successful management of plant diseases. The deployment of this technology aids in
minimizing crop losses and boosting yield quality, contributing to more sustainable
farming practices. Early and accurate disease diagnosis with this methodology not only
mitigates the impact of illnesses on crop productivity but also supports prompt action
lowering the need for expensive chemical treatments and limiting environmental damage.
This study's conclusions underline the promise of deep learning in changing agricultural
diagnostics providing a scalable and efficient approach to strengthening global food
security. The integration of such modern technologies into farming methods is crucial for
promoting sustainable agriculture, enhancing crop resilience and ensuring economic
stability for farming communities globally
