LiveStock Disease Predictor

Website Output
Website Input

🧠 Inspiration Veterinarians and farmers often struggle to quickly diagnose diseases in animals due to limited access to expert healthcare tools in rural areas. Inspired by the rising need for accessible animal healthcare, we built a platform that predicts potential diseases using just a few key inputs: age, type of animal, image, and visible symptoms.

We envisioned a simple, lightweight assistant that leverages AI and image processing to provide early warnings and disease insights, helping reduce animal mortality and improve productivity for livestock owners.

🚀 What We Learned How to build and train a lightweight disease classification model.

Handling and cleaning animal disease datasets, especially multimodal ones (text + image).

Integrating image upload, form input, and ML predictions into a single Flask web app.

Designing a clean and interactive user experience.

🛠️ How We Built It We created a full-stack application using Flask, integrated with HTML5, CSS3, and Bootstrap for the frontend.

The ML model was trained using Scikit-learn and TensorFlow/Keras on a curated dataset of animal images and symptoms.

Users can enter animal age, type, upload an image, and select 3 major symptoms from dropdowns. The backend uses this data to predict the most probable disease.

We implemented image preprocessing (resizing, normalization) and symptom vectorization before feeding data into the model.

Transitions and UI animations were added to create a smooth experience.

🧩 Challenges We Faced Building a generalizable model despite limited animal disease data.

Mapping vague symptoms to concrete disease predictions using AI logic.

Ensuring the frontend and backend communicated smoothly, especially for image handling.

Creating a balance between usability and technical depth in the UI.

Kaggle Notebook :- https://www.kaggle.com/code/virendrabadgotya/livestock-symptoms-random-forest-disease