CropCure: Smart Detection of Potato Diseases Using DL

Inspiration

Farmers who grow potatoes face significant losses every year due to various diseases that affect potato plants. Two of the most common and destructive diseases are early blight and late blight. If these diseases are detected early, farmers can apply appropriate treatments, reducing waste and preventing substantial economic loss. This inspired us to develop a solution that leverages deep learning to help farmers identify these diseases promptly, ensuring healthier crops and better yields

What it does

The model classifies images of potato leaves to detect early blight, late blight, and healthy leaves. By simply uploading a photo of a potato leaf, farmers can quickly determine if their crops are affected by these diseases.

How I built it

I built the model using a deep learning model trained on a dataset of potato leaf images from the Plant Village dataset. The model was developed in Python using TensorFlow and Keras, and trained to distinguish between healthy leaves and those affected by early blight or late blight.

Challenges I ran into

One of the primary challenges I faced was obtaining a sufficient amount of high-quality, labeled data to train the model effectively. Balancing the model to perform well on both early and late blight detection without overfitting was also difficult.

Accomplishments that I am proud of

I’m proud of successfully developing and deploying a deep learning model that can accurately classify potato diseases, which has the potential to make a real-world impact. Additionally, I’m proud of the model's accuracy which is 100% and efficiency, which exceeded my initial expectations.

What I learned

Throughout this project, I gained a deeper understanding of the intricacies involved in training deep learning models, particularly in the context of image classification. I learned how to effectively preprocess and augment data to improve model performance and explored various techniques to prevent overfitting.

What's next for CropCure: Smart Detection of Potato Diseases Using DL

Next, I plan to build a web application that will allow farmers to easily upload images of potato leaves and receive instant disease diagnoses using the model. I’m also exploring the possibility of creating a mobile application to further increase accessibility for farmers in remote areas. Additionally, I aim to enhance the model by expanding its capabilities to recognize more potato diseases and potentially other crops. As I continue to develop this project, I’m eager to collaborate with agricultural experts to ensure that the tool meets the practical needs of farmers and contributes to reducing crop losses.

Built With

• cnn
• keras
• python
• tensorflow