Week 7 - Voronoi 2: Lloyd's Algorithm + Edge Detection

Channel:
Subscribers:
2,640
Published on ● Video Link: https://www.youtube.com/watch?v=-5YcphFnyXM



Duration: 56:41
121 views
5


Today we went over a wide ranging variety of topics, from edge and shape detection to how image compression works, to how to interpolate height data in GIS systems.

The big topics were these:
1) Lloyd's Algorithm for Voronoi is simple, you move a Voronoi seed to the average row of all the rows in its region, and ditto for its column. You then re-run Voronoi and this will tend to smooth out some of the sharper and weirder regions you can get from it.
2) Edge Detection. Basically, you can do simple edge detection by looking at the difference in intensity of one pixel from its neighbor, and marking where there are large discontinuities. For Voronoi, it's as simple as noting where one region transitions into another region, and marking the boundary on both sides.

I want you to get more practice using Git for this assignment since it is the industry standard these days.







Tags:
is50b
wavelet compression
voronoi
lloyd's algorithm
edge detection
object recognition
shape detection
image compression
jpeg compression
git