Surface Crack Detection Using CNN

Developed a Deep Learning image classification model to detect cracks in surface images.

The project used a Convolutional Neural Network (CNN) to classify images into two categories: Positive, indicating the presence of a crack, and Negative, indicating no crack.

The workflow included preparing and organizing the image dataset, resizing images to 224×224 pixels, applying data augmentation, training the CNN model, and evaluating its performance using classification metrics and visual results.

The model was designed to demonstrate how Computer Vision and Deep Learning can be applied to automated surface inspection.

**Tools:** Python, TensorFlow, Keras, NumPy, Matplotlib, Google Colab.