Clustering contours by PCA results

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Clustering contours by PCA results

Maurice
This post has NOT been accepted by the mailing list yet.
Dear community,

using JavaCV 1.0 with OpenCV 3.0.0 i am trying to cluster contours depending on angle and length of vectors generated by PCA. I do not know in advance how manny clusters there are, so i guess K-Means is out of the question. Are there other options?
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Clustering contours by PCA results

keghn
 I am working on something like that.too but in c/c++. Then turn all off the of the contour lines int to descriptor data.
 This information is placed in a data base. Then a search for repeating pattern, or a  preselected pattern , like you would
do in text search.

frieze patterns:
http://www.scientificcomputing.com/articles/2015/09/patterns-are-math-we-love-look

https://en.wikipedia.org/wiki/Frieze_group

https://en.wikipedia.org/wiki/Kolmogorov_complexity


 I guess i could use t-sne or PCA.

Visualizing Data Using t-SNE:
https://www.youtube.com/watch?v=RJVL80Gg3lA

 Is there a existing format for put lines into a data descriptor format?
 I am think it would be a file of pixel positions and the color field they encase. maybe some meta data will
be needed too?


Loading...