Please Hand In to Canvas > Assignments > Assignment1: • A report with your answers and visualizations of the image outputs embedded in the report. Name your report Assignment1Report_AXXX.pdf where AXXX is your student number. • Your source code in the form of a python notebook and intermediate and final image files a .zip named Assignment1Code_AXXX.zip, where AXXX is your student number.
Part 1: Consider the corrupted image (corrupted.png) in Fig. 1(a).
(a) corrupted image (b) recovered enhanced image (c) person crop Figure 1. Image Samples for Part 1. 2. Consider the crop (crop.png) of Fig. 1(c). Are you able to segment the person with the trolley from the background with simple thresholding? If so, which threshold should you use and how did you obtain this threshold? If not, describe why. Support your answer with visualized outputs. (1 mark)
Part 2: Consider the image of the tile mosaic (FranklinMascot.jpg) in Fig. 2(a).
(a) original tile mosaic image (b) red to green conversion (c) sample mask image Figure 2. Image Samples for Part 2
Hints: Consider different variants of superpixelling and Hough transform. Select your search scope and loops wisely for computational efficiency.
