Atlas Honda Inspection Checker

An AI-powered inspection sheet analyzer for Atlas Honda bikes.
Upload or capture a photo of an inspection card, and the app automatically extracts:

• Bike model
• Frame number
• Defected part names (based on tick marks)
• Defects observed from the handwritten confirmation section

This project uses Google Gemini (via LangChain) and Streamlit to create a sleek, interactive web application.

---

🚀 Features

• 📷 Upload or capture inspection card images
• 🤖 AI-powered OCR using Google Gemini
• 🛠️ Extracts structured inspection data
• 🎨 Custom Streamlit UI with animations and branding
• ⚡ Real-time analysis with formatted results

---

🛠️ Tech Stack

• Python 3.10+
• Streamlit
• LangChain
• Google Gemini AI
• dotenv

---

⚙️ Installation & Setup

1️⃣ Clone the repository

git clone https://github.com/Arsalan692/Atlas-Honda-Inspection-Checker
cd Atlas-Honda-Inspection-Checker

2️⃣ Create a virtual environment (optional but recommended)

python -m venv venv
source venv/bin/activate   # For Linux/Mac
venv\Scripts\activate      # For Windows

3️⃣ Install dependencies

pip install -r requirements.txt

4️⃣ Set up API key

🔑 Setting Up Google Gemini API Key

• Go to Google AI Studio: https://aistudio.google.com
• Sign in with your Google Account.
• Click Get API Key from the sidebar.
• Create a new API key and copy it.

5️⃣ Add API Key

Create a .env file in the project root and add your keys:

# GOOGLE API KEY
GOOGLE_API_KEY=your_api_key_here

6️⃣ Run the app

streamlit run Inspection_checker_sheet.py

---

📸 Demo Screenshots