ECE 470: Introduction to Robotics Homework 6
Question 1.
A region of interest (ROI) from row 181 to 200 and column 101 to 120 of an original image I
is specified as ROI_01= I (181:200, 101:120) to encompass a window in the scene with an
array of intensity values as shown in Fig. 1.
Fig. 1
a) For a 5x5 sub-region ROI_02 specified as shown in Fig 1.
i. Write down the expressions in (A) and (B) indexing from image I. (2 point)
ii. Tabulate the available intensity values and their respective occurrence in ROI_2.
Plot the histogram (4 points)
iii. Perform histogram equalization and plot the new histogram (6 points)
iv. Comment on the histograms produced in (ii) and (iii) (2 points)
b) Determine the transformed intensity value of ROI_02 (2,2) if a 3 x 3 mean filter kernel
is applied to the second ROI. (1 points).
Question 2
The image, I is being applied with filters as described below:
F1: Mean filter with kernel size of 15 x 15
F2: Gaussian filter with kernel size 15 x 15 and standard deviation, σ=1
F3: Gaussian filter with kernel size 15 x 15 and standard deviation, σ=3
a) Match the filter with their respective output images in Fig. 2. (3 points)
Fig. 2
b) Match the output images (A) and (B) to the associated edge detection methods
performed, namely, (I) Canny edge detection and (II) Sobel edge detector, assuming the
original image is a noisy image as depicted in Fig. 3. (2 points)
Fig. 3
