Inspiration
Our project was inspired by a personal experience shared by our team member Karan. His mother, an avid gardener, often asked him to use Google Lens to identify diseases in her plants. One day, she discovered that the apples on her apple tree had rotted, leading to both surprise and frustration. This moment highlighted the challenges gardeners face when trying to detect plant diseases early. Motivated by her experience, we decided to create a solution that simplifies the process of identifying plant diseases. Our web application leverages advanced AI technology to analyze leaf images and provide instant feedback, ensuring that gardeners can protect their plants and avoid unpleasant surprises like rotten fruit.
What it does
Our project is a web application designed to assist gardeners and farmers in identifying plant diseases through the analysis of leaf images. By utilizing advanced artificial intelligence techniques, specifically a convolutional neural network (CNN), the application can accurately detect signs of disease and provide detailed feedback.
Users simply upload a photo of a leaf, and the AI processes the image to identify any potential infections. The application not only highlights the disease but also offers actionable recommendations for treatment and prevention. Additionally, a chatbot powered by Google Gemini enhances the user experience by answering questions and providing guidance on plant care.
How we built it
The project was built by first collecting a diverse dataset of leaf images for training a convolutional neural network (CNN) using TensorFlow and Keras. We developed the CNN architecture, tuning hyperparameters to optimize accuracy. A user-friendly web application was created using HTML, CSS, and Flask for the back-end, allowing users to upload images and receive disease predictions. AI tools like ChatGPT and Amazon CodeWhisperer assisted in brainstorming, debugging, and documentation, while Google Gemini powered a chatbot feature for real-time support. After extensive testing, the web app was deployed on a cloud platform, making it accessible to users for easy plant disease detection.
Dataset we used: https://www.kaggle.com/datasets/vipoooool/new-plant-diseases-dataset/data
Challenges we ran into
We faced several challenges during development, particularly since it was our first experience building a web application with a Python back-end. Connecting the AI model written in Python to the front end created with HTML and CSS proved to be a learning curve. It took us time to understand the integration process, troubleshoot issues, and ensure smooth communication between the front-end user interface and the back-end server.
Log in or sign up for Devpost to join the conversation.