I am trying to understand an algorith 2D peak finding

Here is the presentation pdf. I really don’t understand how algorithm work here. . (Page 21/23 to 23/23).

Divide & Conquer #2

Can you narrow down what part is confusing? It’s a big ask for someone to try to re-describe the whole thing.

Okay so the question is Find a peak in 2D array .
In the slide, they have described a method as divide and conquer 2

What does this mean:

" Look at boundary,
center row, and center
column (window)
• Find global max within
• If it’s a peak: return it
• Else:
– Find larger neighbor
– Can’t be in window
– Recurse in quadrant,
including green boundary"

Find global max within?? Within what ?? the center row or center column or where ?.

I’m very interested in this answer. I have no idea what the answer is; but my guess is the goal break it down and work “center out”. edit guess that it’s about making smaller 3x3 areas, but again just assuming /

Without listening to the lecture this PowerPoint is sub-par in my opinion. The grid keeps between slides and I can’t follow along with the logic/example. And there is no result or answer. Maybe it’s a “high level” idea based on the last slide when you use it again “in Module 4”. But if you got a homework practice problem on it; good luck :slight_smile:

1 Like

Hello DVGY.

I am not sure about this, because that PDF does not seem like was originally in English.

  1. The Window is the green portion on page 20.
  2. When it says “global max”, it means local max (seems like an error).
  3. You are supposed to find the max within the window.
  4. If it is not a peak (in this case, it is not, because it is 7), then make a sub-window out of the blue portion, but taking the green around it as the new boundary.
  5. Repeat analysis as necessary.

I hope this helps.

1 Like