/Tire-Texture-Classifier

Explore the implementation of CNN algorithm on the tires dataset to check how it recognises and classifies the texture of tires based on its condition in 2 classes

Primary LanguageJupyter Notebook

Tire Texture Classifier

This project involves training a neural network model to classify tire textures as either "Normal Tire" or "Cracked Tire." The model is built using the Keras library and is trained on a dataset of tire texture images. The trained model is then incorporated into a simple GUI application using Tkinter for image classification.

Files Overview

1. model.ipynb

  • Jupyter Notebook file containing the code for data analysis, visualization, neural network model creation, training, and evaluation.
  • Utilizes the Keras library for building a Convolutional Neural Network (CNN).
  • Implements data augmentation techniques for training dataset improvement.
  • Saves the trained model as 'model3.hdf5'.
  • Evaluates the model's performance on the testing dataset.

2. app.py

  • Python script for a GUI application using Tkinter.
  • Loads the trained model ('model3.hdf5') and defines classes for classification.
  • Allows users to upload an image for classification using the trained model.
  • Displays the selected image and the predicted class (Normal Tire or Cracked Tire).

3. background_image.jpg

  • Placeholder background image for the Tkinter GUI.
  • Replace with the actual path to your preferred background image.

4. upload.png

  • Placeholder upload icon for the Tkinter GUI button.
  • Replace with the actual path to your preferred upload icon.

Running the Application

  1. Install Dependencies pip install tensorflow pillow matplotlib keras scikit-learn