Timeline for Multispectral image segmentation for natural-resources applications using R
Current License: CC BY-SA 3.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 17, 2018 at 15:07 | comment | added | Ka_Papa | @Aaron I already have experience with python and GRASS but I wanted to see how far I can push R. Anyway I'll see what I can find, thanks | |
| Apr 17, 2018 at 13:56 | comment | added | Aaron♦ | @GeorgeNostradamos Not for R. However, Python has a mean shift segmentation that utilizes multiple bands. I believe that is in the scikit-learn package. | |
| Apr 17, 2018 at 9:30 | comment | added | Ka_Papa | Did you find an answer? | |
| May 23, 2016 at 12:18 | answer | added | Tom Higginbottom | timeline score: 3 | |
| May 18, 2016 at 7:15 | comment | added | danclewley | If it doesn't have to be R (just something open source) you could use RSGISLib for the segmentation and apply Random Forests using scikit-learn. There is an example I wrote on how to do this here. spectraldifferences.wordpress.com/2014/09/07/…. You could also use RSGISLib to segment and attribute objects then import into R as a text file for the classification. | |
| S Dec 26, 2015 at 8:04 | history | bounty ended | CommunityBot | ||
| S Dec 26, 2015 at 8:04 | history | notice removed | CommunityBot | ||
| Dec 18, 2015 at 6:51 | history | edited | radouxju | CC BY-SA 3.0 |
added 78 characters in body
|
| S Dec 18, 2015 at 6:49 | history | bounty started | radouxju | ||
| S Dec 18, 2015 at 6:49 | history | notice added | radouxju | Draw attention | |
| Jan 28, 2015 at 11:23 | comment | added | pixelmitch | In the example you link, they use spectral clustering. Spectral clustering in principle can be applied to n-dimensional data as well, as it simply decomposes a similarity matrix that can be built for any type of data. The principle of spectral clustering is to embed data in a lower dimensional space and perform clustering in that space (e.g. with kmeans). I'm not very familiar with R, but I'm sure it would only take few lines of code (at least in matlab, it does take few lines). If not, I'm sure there are free implementations of other approaches, such as mean-shift segmentation. | |
| Dec 13, 2014 at 5:13 | history | tweeted | twitter.com/#!/StackGIS/status/543634853775028224 | ||
| Aug 6, 2014 at 13:04 | history | edited | Aaron♦ | CC BY-SA 3.0 |
Rearranged post for clarity
|
| S Aug 5, 2014 at 22:50 | history | suggested | Leasye | CC BY-SA 3.0 |
add spaces
|
| Aug 5, 2014 at 22:35 | review | Suggested edits | |||
| S Aug 5, 2014 at 22:50 | |||||
| Aug 5, 2014 at 22:31 | history | asked | Aaron♦ | CC BY-SA 3.0 |