A collection of Python notebooks demonstrating fundamental computer vision filters (Edge Detection, Blurring, and Sharpening) using TensorFlow and Keras 2D convolutions.
This repository contains interactive Jupyter Notebooks designed to visualize how Convolutional Neural Networks (CNNs) process images. Using TensorFlow and Keras, it applies various 3x3 mathematical kernels—such as Horizontal/Vertical Edge Detection, Gaussian Blurring, and Edge Enhancement—to a single-channel grayscale image. It serves as a practical, hands-on introduction to image filtering and feature extraction in computer vision.
1_Edge_Detection_Filters.ipynb: Demonstrates horizontal, vertical, diagonal, and Laplacian edge detection filters.2_Blur_and_Sharpen_Filters.ipynb: Demonstrates box blur, Gaussian blur, sharpening, and edge enhancement filters.Test Image/: Contains the base image used for applying filters.Result/: Contains output composite images generated from the notebooks.
You can run these notebooks locally or upload them to Google Colab. If running in Colab, the notebooks are configured to fallback to a sample image if the local test image is unavailable.
