0
$\begingroup$

The problem is to determine number of objects in the area defined by polygon of arbitrary shape. Suppose trees are planted randomly in a rectangular area. The trees are being eaten by some bug and after some time die. Some trees do survive. Both surviving and dead trees form the clusters. In general there are more than 2 clusters. The question is 1) how to detect boundaries between clusters of surviving and dead trees, 2) count the number of trees within each cluster, 3) determine the area of each cluster. I am determining the boundaries using R function boundaries from raster library but got stuck with 2) and 3). Any help is appreciated.

$\endgroup$
5
  • $\begingroup$ What data do you have or propose to collect? $\endgroup$
    – whuber
    Jan 8 '20 at 20:01
  • $\begingroup$ ...or is that part of the question also? $\endgroup$
    – Ben
    Jan 8 '20 at 20:36
  • $\begingroup$ The bug is EAB and data is collected on ash trees. $\endgroup$
    – TFbie
    Jan 9 '20 at 12:56
  • $\begingroup$ Yes, but what kind of data and in what form? Satellite images? Samples? Transects? Quadrat counts? $\endgroup$
    – whuber
    Jan 9 '20 at 20:33
  • $\begingroup$ year, lat/long of tree, alive/dead status (determined by visual inspection). Data is for several years so one can see how infestation progresses. There are about 400 trees. It is in spreadsheet. $\endgroup$
    – TFbie
    Jan 9 '20 at 20:50
0
$\begingroup$

I think the answer is triangulation. Triangulate all trees and then break the edges where one vertex is dead tree and another vertex is surviving tree.

$\endgroup$
1
  • $\begingroup$ Yes, Delauney triangulation did the job. So simple... $\endgroup$
    – TFbie
    Jan 17 '20 at 17:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.