Yannick I am so appreciative of your recommendations related to tutorials and getting to know these parameters re the OTB.
I will follow your instructions and will keep you posted as per my success with this OS software. I'm used to using ENVI and ERDAS Imagine but am interested in the OS platform. Thankyou most kindly for your response. ~Natasha On Wed, Feb 28, 2018 at 11:05 AM, Tanguy Yannick <[email protected]> wrote: > Hi Natacha, > > > > These parameters can be quite hard to choose and it depends on your images. > > - Spatial radius is used for the segmentation step : it’s a > value in pixel, and if you choose a too big value, the algorithm will spend > a lot of time to analyze and segment the image. > > - Mode convergence and nb of iterations both help to stop the > algorithm > > - Maybe if your image has a lot of small regions, you can > decrease a little bit the minimum region size (value in pixel). > > > > There is an exercise about segmentation in the OTB training course [1]. > > I suggest you have a look at that exercise : it’s a nice way to better > understand OTB and how to use this application. > > > > If you want to analyze the intermediate results, you can also use the > different algorithms separately : > > otbcli_MeanShiftSmoothing, otbcli_LSMSSegmentation, [ > otbcli_LSMSSmallRegionsMerging ], otbcli_LSMSVectorization > > > > I think this can help you find the better parameters that fit your needs. > > > > Hope it can help ! > > Best regards, > > > > Yannick > > > > [1] see https://www.orfeo-toolbox.org/documentation/ > Training material > > There are “org” files under our git repository. These org files can be > converted in pdf files, or read with a simple text editor. > > > > > > *De :* [email protected] [mailto:[email protected]] *De > la part de* Natasha gregory-michelman > *Envoyé :* mercredi 28 février 2018 03:43 > *À :* otb-users <[email protected]> > *Objet :* [otb-users] New to OTB Segmentation Application; how to > decipher parameters in segmentation > > > > I am very new to the Orfeo Toolbox and am having trouble deciphering how > to successfully segment a multispectral image.have all the tools loaded > into QGIS 2.18 and have evaluated them in the Monteverdi application as > well. > > I also have the OTB cookbook up and have read up on the Segmentation > application in chapter 4 page 32 which doesn't indicate what values are > appropriate to enter into the Parameters window to get a valid segmentation > output either in raster or vector format. > > (using mean shift here) > > > > 1. What are the numerical parameters that are recommended for a MS image > with a 4 band spectral range (to make it simple). > > a) Spatial Radius is default 5_how would I know what to enter without > seeing the range radius and how to change it to the appropriate numeric > value? > > b) Mode convergence threshold is 0.1000000 What does this apply to? > How should I amend it? > > c) Maximum numbers of iterations (100) this is clear > > d) Minimum region size (100) this seems like a very important > component parameter but if I am working with a 10 m pixel res I was > thinking this would be a good way to start and have gotten a black image as > an end product. > > > > > > 2. Is there a way to see the segmentation process as it is being applied > when instituting these parameters? > > > > Again_I'm new to the OTB and am anxious to learn so if anyone has some > good information to pass along, I would welcome it. > > > > I'm trying to learn some code along the way and this is a good platform to > realize it. > > > > Thanks for any information. > > > > ~Natasha > > > > [image: Image supprimée par l'expéditeur.] > <https://lh3.googleusercontent.com/-5ryp8T4HUWI/WpYWsASCc7I/AAAAAAAAAgY/RqGWL6wuzbQRj3xb2iHq8lqDifzA57mQQCLcBGAs/s1600/orfeo_TBX_parameters.jpg> > > -- > -- > Check the OTB FAQ at > http://www.orfeo-toolbox.org/FAQ.html > > You received this message because you are subscribed to the Google > Groups "otb-users" group. > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected] > For more options, visit this group at > http://groups.google.com/group/otb-users?hl=en > --- > You received this message because you are subscribed to the Google Groups > "otb-users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > > -- > -- > Check the OTB FAQ at > http://www.orfeo-toolbox.org/FAQ.html > > You received this message because you are subscribed to the Google > Groups "otb-users" group. > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected] > For more options, visit this group at > http://groups.google.com/group/otb-users?hl=en > --- > You received this message because you are subscribed to a topic in the > Google Groups "otb-users" group. > To unsubscribe from this topic, visit https://groups.google.com/d/ > topic/otb-users/QhBgefxejzU/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- -- Check the OTB FAQ at http://www.orfeo-toolbox.org/FAQ.html You received this message because you are subscribed to the Google Groups "otb-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/otb-users?hl=en --- You received this message because you are subscribed to the Google Groups "otb-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
