OTB algorithms are in OTB_installation_path/bin. So, as you can use them in command line (see more in documentation), the same way could be used trough R by system function.
You can create your owns functions to work with OTB in R. An example for Mean Shift Image Segmentation:
meanshift.segm <- function(otb.path = "", raster.in = "", out.path = "", name ...
This issue is caused by the workflow intended by OTB developers. Pan-sharpening in OTB is meant to be performed using console command BundleToPerfectSensor: otbcli_BundleToPerfectSensor -inp pan_image -inxs xs_image -out output_image.
As described corresponding section of the manual - if one wants to perform pansharpening using GUI (Monteverdi), one have to ...
The Edge Extraction tool may have been moved to the Orfeo Toolbox as this focuses on image analysis. The Orfeo Toolbox contains a number of edge extraction tools using various methods such as:
The OTB provider was removed from QGIS 3.0. See the QGIS 3.0.0 changelog: http://changelog.qgis.org/en/qgis/version/3.0.0/#removed-otb-lidartools-providers-from-processing
As long as nobody makes a QGIS plugin to add OTB as a processing provider, it isn't possible to use OTB in QGIS 3.0.
To better understand, I have done a similar segmentation in GRASS GIS 7 using the new i.segment. In my opinion these lines appear where the image data were mosaiked due to a non-perfect histogram matching (or whatever) being used.
In short: Orfeo or any other segmentation software may deliver better results when the initial mosaiking is improved to avoid ...
It may also depend on the algorithm. For example, the Ordinary Kriging from SAGA returns two outputs (VARIANCE and PREDICTION), but other SAGA algorithm outputs are called OUTPUT. So, you cannot just use a single output name for all algorithms in a provider.
However, there is a way to get output parameter names from any Processing algorithm. We can define ...
You are right that orfeo toolbox is not yet available in the 64 bit Version of OSGeo4w installer. You can compare the package listings here:
The 64-bit version is still experimental, with packages missing or not working as expected.
However, it is ...
there is a thread in the otb user mailing list which give a workaround for this issue:
It is related to the fact that qgis is doing something specific when the input image does not have an origin. The workaround is to make an extract from ...
Sextante is now known as Processing and is a core plugin (no need to install). OTB, if using Kyngchaos.com QGIS installer, is included inside the QGIS.app bundle.
Processing should be able to access the OTB applications within the QGIS.app bundle, i.e. with the Kyngchaos.com installer Processing should just work out-of-the-box.
Also bundled within the ...
I've work with both OTB and SAGA but I would advise on Ubuntu to install both libraries using the ubuntugis-unstable PPA:
It should allow you to install last stable release of both otb and saga and also qgis using the same gdal version.
Take care to uninstall previous versions on your ...
You might consider using GRASS (which interfaces with QGIS) for this process. See this link for more information: http://grasswiki.osgeo.org/wiki/LANDSAT
The answer provided by @markusN on this question provides additional information and examples of what GRASS can be used for with Landsat-8.
It may have to do with some conflict between the two versions of SAGA (2.0.8 and 2.1). When unchecking the box 'Enable SAGA 2.0.8 compatibility' in Settings and configuration, it solved the problem for me. I´m using QGIS 2.0.1 Dufour with SAGA 2.1 on Windows 7 64 bit. Hope this will help.
I found a tutorial here
But it is not that helpful, as when I am in the preparation of reference data (join attributes by location), it results in shapefile, and there is no XML file. Meanwhile, the next step requires an XML file.
In the end, I classified them based on ruleset like eCognition, but have to write the script down in field ...
Most segmentation processes require a lot of processing time, but if you need 5 days for 0.6km² you must have a very high spatial resolution image. By resampling to a coarser pixel size before you run the segmentation algorithm, you should be able to run a lot faster. For trees, 25 cm or even 50 cm is often enough (depending on the tree type).
Ok, I am responding my own question because I have solved it for my own purpose. It doesn't mean that there could not be a better answer.
It seems to be due to the version. I downgrade to the LTR version of QGIS, and the Orfeo Toolbox appear in the processing toolbox!!
My steps where:
sudo apt-get remove qgis python-qgis
sudo apt-get ...
What parameters did you try for the segmentation?
The OTB segmentation module got 2 modes, a raster mode which outputs a image of label and a vector mode which works on larger image and include a tilling scheme. This last mode output the segmentation results in vector mode.
Have a look to this post which describe the application:
note that next OTB release (3.20) will provide a new segmentation workflow based on the mean-shift algorithm which allows to perform tile-wise segmentation of very large image with theoretical guarantees of getting identical results to those without tiling.
You can find a description of the workflow (different from the one cited above) here.enter link ...
It seems you are encountering this bug and the proposed workaround is to work in WGS84.
for using OTB, a workaround is to install the stand alone Monteverdi software (directly based on OTB), also available from OSGEO4W. Using the task specific OTB applications could also be more "robust". (see the cookbook for more details)
Before actually using OTB from a Python instance on your computer you have to make sure you have access to the relevant libraries. ArcGIS installs Python in a separate folder from that of OSGeo4W. So, if you are planning on using both ArcPy and OTB you must look closely at the paths and locations of your libraries.
For using OTB from within Python when ...
Installing QGIS Dufour standalone, I have not found any path with orfeotoolbox files. Just because you see the menu entries in the plugin, it does not mean they are already installed.
The otb website recommends to install it via OSGEO4W installer. So even if you want to use QGIS standalone, you are encouraged to install otb with the OSGeo4W system, and add ...
approximate vector results of the mean-shift on full image using the Segmentation application with the vector mode:
Or you can produce an exact segmentation (raster) of the full image following the LSMS workflow which require to use 3 different applications in a row:
I think the issue is in the following line:
outputImages = os.path.join(root, tifFile) + '_concat.tif'
which returns something like: path/to/the/tifFile.tif_concat.tif
Instead, you need to strip the extension from the original file:
outputImages = os.path.join(root, tifFile)[:-4] + '_concat.tif'
EDIT based on the comment:
It looks you need to provide a ...
I think that OTB was not able to read samples class in the input polygons
Labels should be in the shapefile in a column attribute. By default OTB is looking for a column name 'Class' but you can change this value in the SVM trainer application.
Use Orfeo Toolbox > Image Manipulation > Split Image. This splits an n-band multiband image into n images. The output filename is a prefix and the channels (bands) are numbers accordingly e.g. if filename = 'myImage.tif' then the first band will be saved as 'myImage_0.tif' and the second 'myImage_1.tif' and so on. See here for a more detailed ...