Leather Quality Classifier Web Application

This is a web application that uses a trained InceptionV3 model to classify leather quality. The model can classify leather images into 4 different categories.

Features

• Modern, responsive web interface
• Drag and drop image upload
• Real-time image preview
• Classification results with confidence scores
• Support for common image formats (JPG, PNG)

Requirements

• Python 3.7+
• TensorFlow 2.x
• Flask
• Other dependencies listed in requirements.txt

Setup

1. Clone the repository:

git clone <repository-url>
cd leather-classifier

2. Create a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Make sure you have the model file:

• The model file inceptionNetV3_100eVAL_16b_v2_model.h5 should be in the root directory
• This file contains the trained InceptionV3 model for leather classification

Running the Application

1. Start the Flask server:

python app.py

2. Open a web browser and navigate to:

http://localhost:5000

Simplest API Deployment (FastAPI)

For a minimal prediction service (no UI/login) use simple_api.py:

1. Install requirements (FastAPI/uvicorn already listed):

   pip install -r requirements.txt

2. Ensure the Inception .h5 model is available and set the path (defaults to inception.h5 in repo root):

   set MODEL_INCEPTION_PATH=inception.h5  # Windows PowerShell
   export MODEL_INCEPTION_PATH=inception.h5  # Unix shells

3. Start the API with uvicorn:

   uvicorn simple_api:app --host 0.0.0.0 --port 8000

4. Send predictions via HTTP:

   curl -X POST "http://localhost:8000/predict" \
        -H "Content-Type: multipart/form-data" \
        -F "file=@path/to/leather.jpg"

This spins up a CORS-enabled FastAPI server that loads the model once and returns JSON predictions.

Legacy implementation

The original Flask UI, templates, static assets, and alternate FastAPI stack were moved into legacy/ for safekeeping:

• legacy/flask_app/ – original app.py, templates, and static files
• legacy/fullstack_api/ – earlier FastAPI service with database helpers
• legacy/docs/ – previous Elastic Beanstalk runbook
• legacy/models/ – unused/alternate .h5 artifacts (e.g., AlexNet)

Only simple_api.py is required for the current proof-of-concept deployment.

Usage

1. Click the upload area or drag and drop an image file
2. Preview the image
3. Click "Classify Leather" to get the prediction
4. View the classification result and confidence score

Model Information

• Base Model: InceptionV3
• Input Size: 224x224 pixels
• Classes: 4 leather quality categories
• Training Accuracy: >97%
• Validation Accuracy: >97%

Technical Details

• Frontend: HTML5, CSS3, JavaScript
• Backend: Flask (Python)
• Deep Learning: TensorFlow/Keras
• UI Framework: Bootstrap 5

License

[Your License Information]