Computer Vision Lab 1

For National University of Singapore (NUS) module CS4243 Computer Vision. Completed as pairwork.

All programs are written in Python, making use of numpy, cv2 and matplotlib libraries, and run on Jupyter Notebook.

Learning Objectives & Implemented Functions:

• Basic image pre-processing, such as padding zeroes and estimating gradient map from grayscale image (convolution with Sobel filters)
• Normalised cross-correlation for both grayscale and RGB images (with multiple implementations, using element-wise multiplication and matrix multiplication)
• Mean-subtracted cross-correlation
• Non-maximum suppression to obtain local maxima
• Template matching - considering different template patterns and sizes, limitations etc.

Note: the exact questions can be found in the PDF lab sheet. To visualise the results of the above functions, the output obtained from running on sample input can be found in the ipynb file.

Link to Lab 2 here