Computer Vision help

[vernal whale]

So I am currently studying for my Computer Vision mid-term examination and I am practicing it from previous years exam papers. There are 3 questions in the paper. I have difficulties in doing questions 2 and 3.
https://drive.google.com/drive/folders/1wwCAX0Hq0p4SCd2k0OWkSjUQdWOMC13V?usp=sharing
The exam paper is the one which name is "UTS_ComVis_2020.pdf".
Are there anyone who are willing to help me?
Any help will be greatly appreciated.

[shadow moth]

According to the guidelines of the forum #📄・posting-guidelines we cannot help you do academic assignments. Having said that, Can you elaborate what you have tried doing and explain what difficulties your are facing ? May be we can help in resolving those!

[vernal whale]

the code is in the link, the ones which names are "LATIHAN-UTS-CV2" and "LATIHAN-UTS-CV3"

[vernal whale]

thanks but the result is still not that good tho

[shadow moth]

The idea 💡 I used was to remove the foreground and then use canny edge and houghlines to get the edges.

You could try to tuning the threshold for getting better results

It seems much better approach then just masking the foreground as you had in your earlier code

[vernal whale]

ah ok

[vernal whale]

which threshold tho? the threshold value in cv2.threshold or the threshold value in cv2.HoughLinesP?

[shadow moth]

We are using threshold values in couple of functions. Try with HoughLinesP first, then try with others. See which one works best. You can also try evaluating a mean value that works for 3 images