Q1: For the following magnitude and orientation images, estimate the edge image value after nonmaximum suppression for the pixels located at (3,2) and (4,5).
(a) magnitude (b) orientation
Coding (see CS4243-Tutorial-02.ipynb):
β’ Add Gaussian noise with π=0 and π=20 to the image Nemo.jpeg to produce a noisy image. After adding Gaussian noise, apply a floor and ceiling operation to ensure that all pixels remain within [0-255].
β’ Apply the cv2.Sobel with size=3 to the noisy outputs and visualize the outputs
β’ Smooth the noisy image with a Gaussian filter of size=15 and sigma={1, 3, 9} and compare
β’ Apply the cv2.Sobel with size=3 to the different smoothed images and compare the outputs. Are there any differences in the extracted edges?
β’ Apply built-in cv2.Canny to the image Pole.png, using {50, 100} as {threshold1, threshold2}. Compare these outputs when {80, 100} and {50, 300} as thresholds. Are there any differences in the extracted edges?
Past Quiz Questions:
β’ AY2021_Quiz2: Q1
β’ AY2122_Quiz1: Q5
