2026
Computer vision fundamentals with OpenCV
Four labs on color channels, geometric transforms, filtering (box and Gaussian) and edge detection with a hand-built Sobel operator.
Problem
Understand what each operation does to pixels before moving on to convolutional networks.
Data
Sample photos loaded from /content in Colab; any RGB image works.
Method
Implementations in OpenCV and NumPy, a Gaussian kernel built from its formula, and X/Y gradients, with reasoned answers to each prompt.
Result
A hand-built smoothing and edge-detection pipeline, the groundwork for the EuroSAT CNN project.
Stack
- Python
- OpenCV
- NumPy
- Matplotlib
Context
Course: Image Processing · 2026