Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.soft-sys.math.mathematica > #9850
| From | Szabolcs Horvát <szhorvat@gmail.com> |
|---|---|
| Newsgroups | comp.soft-sys.math.mathematica |
| Subject | Re: 2D Gaussian fit |
| Date | 2012-04-28 09:23 +0000 |
| Organization | Steven M. Christensen and Associates, Inc and MathTensor, Inc. |
| Message-ID | <jngcuo$j5$1@smc.vnet.net> (permalink) |
| References | <jnb4m4$5ic$1@smc.vnet.net> |
On 2012.04.26. 11:31, zhaol01@gmail.com wrote: > How can I fit a spot on an image to a 2D Gaussian function? > I have a Tiff image file, on it there's an almost round spot, e.g. laser profile. > Is there a way in mathematica to 2D Gaussian fit the spot? > Hello, The best way to do this would depend on the kinds of images you have (i.e. how noisy they are, how many spots are there, etc.) The functions here could be useful to you: http://reference.wolfram.com/mathematica/guide/SegmentationAnalysis.html For example ComponentMeasurements will help measure the spot and also identify its centre. Then you could cut the spot from the image and convert it to a matrix so you can use fitting methods such as those implemented in NonlinearModelFit. I recommend looking at this blog post as well: http://blog.wolfram.com/2012/01/04/how-to-count-cells-annihilate-sailboats-and-warp-the-mona-lisa/ -- Szabolcs Horvát Visit Mathematica.SE: http://mathematica.stackexchange.com/
Back to comp.soft-sys.math.mathematica | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
2D Gaussian fit zhaol01@gmail.com - 2012-04-26 09:31 +0000 Re: 2D Gaussian fit Szabolcs Horvát <szhorvat@gmail.com> - 2012-04-28 09:23 +0000 Re: 2D Gaussian fit FairView <zhaol01@gmail.com> - 2012-05-01 09:23 +0000
csiph-web