Computer Vision 2022
代写assignment | lab代做 | Computer Vision代写 | CV代写 – 这个项目是Computer Vision代写的代写题目
Workshop 1
Preliminary
Workshop sessions in Computer Vision are not assessed. Each session will have a few ques- tions aimed at starting discussion between you and the tutor, but you are free to come along with your own questions or points! Participation in these sessions is for your own benefit. The aim of this session is to make sure everyone is familiar with the basics of image filtering and making progress on their first assignment. There is no coding required, but if you have a specific question about some code you are writing it would help to bring your laptop along.
Question 1 Point processes
In the green screen technique often used for visual effects (and Zoom meetings!), an actor works in front of a green backdrop which is later replaced by a rendered scene. What point process would you apply to separate the pixels on the background from the pixels on the actor? What problems might arise?
Figure 1: Cheesy actor in front of green screen.
Question 2 Image filters
In the green screen example above, errors in individual pixels can lead to speckle type noise in the foreground and background regions. How could you remove this noise with a filter? Hint: have a look at the section on morphology filters in the textbook.
Case Study
- Get the humans mask with Google co lab : https://colab.research.google.com/ drive/13x-Fqu2g1iDKO-BieA_75L_TAlVnCLl4?usp=sharing
1
assignment 1 check-in
How are you finding assignment 1 so far? Any remaining time in the session can be used for general discussion of your progress so far, any stumbling blocks and so on.
Yifan Liu
March 2022