With the Pleiades launch a few weeks ahead, there has been a lot of ongoing work from the OTB team to prepare for these new data. Since a lot of discussions happened off the list, either with phone meetings, or on other mailing lists (we will see later on why), and since we now have a comprehensive knowledge of what the support of Pleiades images in OTB will be, it is about time we explain it to users and developers.

Reminder on Pleiades images

We will start with a reminder : Pleiades images will be available in Jpeg2000 format allowing a high compression rate, tiled in 2048×2048 pixels, so a standard Pleiades image will contain a few hundreds of tiles. A typical Pleiades image size is 40 000 x 40 000 pixels (corresponding to a 20 km by 20 km area), and one of the standard product is a pan-sharpened one, which is a merge of the high resolution panchromatic and lower resolution multispectral imagery to create a single high resolution color image. These images would be very heavy without Jpeg2000 compression : for instance, if the Jpeg2000 file weights 1.7 Go, the decompressed file weights 7.3 Go. To decode those JPEG2000 files, there are a few commercial libraries, the most popular beeing Kakadu. There is also some open-source alternatives, among which OpenJPEG seems to be the most advanced. Of course, for Pleiades support in OTB, we need an open-source solution, even if GDAL can be compiled with a driver based on Kakadu (has to be the commercial version, not the trial one available for free under restrictions of use).

Open-source JPEG 2000 implementation

Back in 2007, the OTB people here at CNES already spotted OpenJPEG as the best open-source bet for JPEG2000 support in OTB. Since the library was missing some important features (like partial decoding and MCT support), they set-up a CNES contract with the CS Company in order to add these features to OpenJPEG. But when this contract was over, the new features were not integrated in OpenJPEG trunk but left on a side branch (the so-called v2), because nobody in OpenJPEG community had in-dept knowledge of what had been done in this new version. The development of the trunk went on with bug-fixes and enhancements, while the v2 branch did not evolve. In the meantime, we started a driver in OTB based on the v2 OpenJPEG version, but this was not a sustainable option, because the v2 was barely maintained. Thanks to its more advanced functionnalities, the v2 branch of OpenJPEG also received interest from other FOSS projects in need for advanced decoding capabilities, and got integrated in at least ITK, GDCM and GDAL.

Four months ago, we agreed with the OpenJPEG community that we needed to get the v2 merged with the OpenJPEG trunk. Mickaël Savinaud from the CS OTB team got involved into OpenJPEG development to get the merge between v2 and trunk done, and we now have a full-featured version of OpenJPEG in trunk thanks to his great work and to the support of the OpenJPEG community.

Now, it is time to face the truth : even with this new version, in terms of decoding performances, this state-of-the-art open-source software is way worse than Kakadu : decoding one tile is way slower than with Kakadu, and one tile of 2048×2048 pixels is the atomic unit for decoding (i.e. if you need one pixel, you will still need to first decode the whole tile).

We looked into optimization : reducing file seeking, implementing some macroscopic code optimisation (mainly avoiding pessimisation) of the Tier1 part which is clearly identify as the performances bottleneck (see figure below with profiling reports using kcachegrind). We made some clear progress, but the JPEG2000 standard is complex, and without knowledge of the big picture, we could not gain a lot.

KCachegrind representation of time percentage per functions

What does it mean for Pleiades data ?

It means that simply decoding a full Pleiades image at full resolution will take about 20 minutes on a decent i5 CPU with 4 Go of RAM, while the same image in standard TIF format would have taken 8 minutes and Kakadu takes only 4 minutes. Any OTB processing pipeline streaming the whole image will be limited by this decompression time. And we are not even talking of sub-sampling or estimating statistics at some point in the pipeline.

Of course, we could tell our users to buy a Kakadu licence and compile GDAL with the Kakadu driver enabled, but we can not even tell them how much it will cost, and this is clearly in contradiction with the open-source philosophy of OTB. Still, this remains an option to get high performances JPEG2000 support in OTB, but you are on your own if you choose this solution.

Now, what will it be possible to do using Pleiades data with the open-source solution OpenJPEG in OTB ?

• On OTB side: No specific handling will be added. It is the responsibility of the developer to know that the JPEG2000 driver is not as fast as the other drivers. The developer has the possibility to easily select the appropriate resolution and accordingly set the size streamed region (to reduce the number of decoding operations).  Of course, metadata related to geometric and radiometric calibration will be supported through OSSIM. Moreover, we will support the GMLJP2 box included in the Pleiades to support geo-information in case of ortho-image products.

• On applications side:

• We will provide an application able to convert a Pleiades image, either full or extract, from any JPEG2000 resolution, into another file format (like TIF for instance). Of course, one will need a lot of disk space to handle the decompressed data. Please note that this is not fully supported by Kakadu trial version, which decodes a maximum of 3 channels (and which use is restricted by the way),

• All applications will support Pleiades images, but it will be highly recommended to first decompress the image to disk for reasonable performances and if more than one processing is to be executed on the image,

• On Monteverdi side:

• Pleiades images will open in a dedicated new type of data in Monteverdi interface, and the user will be able to select the resolution level for decoding,

• The viewer module will accept this new type of data and we will be able to efficiently navigate within Pleiades images : fast quicklook computation due to JPEG2000 capabilities, instant navigation into a 4 tiles squared area thanks to caching, and quick-enough refresh when moving outside this cached area thanks to parallel tiles decoding. We expect the navigation experience to be quite good !

• The extract ROI module will accept this new type of data and allow to select an extract of the image using the fast quicklook ability. Caching (which will decompress image to disk in TIF format) will be recommended (this behaviour might be extended to other modules later).

• The writer module will accept this new type of data for conversion purposes.

• All other modules will not accept this new type of data : to use them, the user will first have to use the extract ROI module which will produce a plain image type accepted by all modules. Caching will be recommended for this purpose.

Closing Thoughts

We can hope that the widening use of JPEG2000 in spatial imagery will encourage the improvement of open-source JPEG2000 libraries performances. When and if this becomes available, you will be able to use those datasets in OTB seamlessly, as any other image.

Until then, we can look at the bright side :

OTB will provide an open-source and free access to Pleiades data for those not owning a commercial image processing software  (please note that Gdal is looking into the new OpenJPEG version and might also support Pleiades through it in a near future).

OTB will provide an efficient solution for Pleiades data decompression and visualization. 

Last, the whole set of OTB processing remains fully available, at the expense of extra decoding time or once your Pleiades data have been decompressed to disk.

This has also been a great experience of collaboration between open-source projects towards a common goal. We would like to thank again the members of the OpenJPEG community, as well as the OTB Team at CS for their great work !

It took him no less than 7 hours in two days to do it, including building the cake pan out of carton pieces. It is needless to say it took us far less time to eat it once cut, and we can assure you that this cake was as tasty as beautiful !

Cheers !

Julien Michel

The idea is more or less the same as for the GSoC : ESA selects open source projects as mentoring organizations, and propose to pay students a very decent amount of money to work on those open source projects during a few months in the summer.

Students have the opportunity to work on fun projects during the summer, and in the end they can add an interesting experience to their resume.

The list of mentoring organizations selected by ESA can be seen here. As you can see, Orfeo Toolbox is selected for this pilot year !

An ideas list for this program has been compiled here, so students can start applying by picking up ideas from there. Of course, any other project which improves the library will be welcome, as long as we have the opportunity to discuss it before the application.

This year being a pilot year, the schedule is very tight, and students must apply before the 27th of July.

Students : we are looking forward to hearing from you soon !

CNES is  happy to announce that the Orfeo ToolBox has entered the release candidate stage. The packages for the stable 3.10 release will be available at the end of June. It will coincide with the fifth anniversary of the Orfeo ToolBox library (the release 1.0 was done the 06/30 2006!).

The packages for OTB (Library 3.10, Applications 3.10, Monteverdi 1.8 and Wrapping 1.6) Release Candidate 1 are available here:

• http://www.orfeo-toolbox.org/packages/OrfeoToolbox-3.10.0-RC1.tgz
• http://www.orfeo-toolbox.org/packages/Orfeo-Applications-3.10.0-RC1.tgz
• http://www.orfeo-toolbox.org/packages/Monteverdi-1.8.0-RC1.tgz
• http://www.orfeo-toolbox.org/packages/OTB-Wrapping-1.6.0-RC1.tgz

For MS-Windows users, there are also a binary auto-install packages available for Monterverdi (here), OTB-Applications (here) and OTB-Wrapping (here) !

If you feel like it, you can download them and test the install procedure. Any feedback will be warmly welcomed! There are lots of new features in this release! The list of new main add-ons is available here.

There are exciting new stuffs in all corners of the library. New functionnalities include among other things :

• A new multi-image supervised classification suite in OTB-Applications described here.
• An application to perform segmentation based on connected components and object based image analysis (see description here)

• A framework for cartographic database creation and validation based on the fusion of features in the framework of Dempster-Shafer evidence theory (see description here)

The Monteverdi application is now composed of 42 modules and there are 42+1 applications in OTB-Applications!

The OTB development team also initiated, in addition to the OTB software guide, a guide for OTB-Applications and Monteverdi, dedicated to non-developers. This guide is composed of a brief tour of OTB-Applications and Monteverdi, followed by a set of recipes to perform usual remote sensing tasks with both tools. A first version is already available here.

Please feel free to send your comments, reviews or contributions!

Best regards,

Manuel GRIZONNET, on behalf of OTB development  team.

The French Space Agency (CNES) in collaboration with the French Institute for Research and Development are helping to the development of the use of remote sensing data by getting access to images (Spot) but also by providing tools to manipulate these data.

In 2009, for the first course in Antananarivo, Monteverdi was developed. The software is based on OTB and gives access to lots of functionalities through an interactive and modular structure. It gets access to a wide variety of technical modules and takes advantage of streaming and multi-threading capabilities of the library. We received a lot of interesting feedback after the course, requests for the implementation of new functionalities grew steadily and so the development of Monteverdi continued during 2010.

On the other hand, Monteverdi is a “young” software which could be improved but it is a step to help the development of a pragmatic and interactive tool to manage the use of remote sensing data. A tutorial was just organised on this subject at the IGARSS conference in 2010 in Honolulu.

The second course in Antananarivo, two weeks ago, was the occasion to capitalize on concepts learnt in 2009 and was also the occasion to present new functionalities developed during the last 12 months. The courses covered a large scope of the main remote sensing processes available in Monteverdi. From basic operations to the practical use of the Change detection framework or of the optical calibration module based on the radiative transfer code 6S.

One more time Monteverdi proves its abilities as a support for technical courses because it can cover a large scope of the principal tasks needed by a user of remote sensing data. From geometric processing like the orthorectification module to Feature extraction, segmentation or supervised classification.

It also appears that some new modules added in 2010 get a lot of success and are a progress in the right direction. For example, the Band Math module which allows to perform generic mathematical operations between image layers.

Moreover, this kind of tutorials are also an excellent way to get immediate feedbacks and reactions about the software and to get a direct interaction with end users:

-Is there a module to perform Object based classification in Monteverdi ( notably for VHR data)? YES, have a look to the Object labeling module

-Is it possible to save/load process chain in Monteverdi? For now no, but we will work on it…

-Is it possible to apply a process to a masked region of the input image?

-Is it possible to remove results from the data tree in Monteverdi? NO (but it is feasible)

The vast majority of students’ questions or problems often point out really interesting feature requests and as a consequence are the beginning of the roadmap of the software.

There are also more general questions about the software:

-How to increase the ergonomy and the user experience of Monteverdi?

-How can we generalize the integration in Monteverdi of existing OTB filters for non developers?

-How to achieve a seamless integration and collaboration between OTB and existing Open Source tools?

One more time the course was also a proof of the value of Open Source principles: for example the combination of Open Data over Madagascar available through the Open Street Map project and open source software able to use it efficiently like Monteverdi.

The idea is not to believe in a “techno-utopia” where information technology could eliminate poverty and inverse difficult solutions. Information technology and Open Source software increase capabilities but are dependent on the fact that they need users which are ready and able to use them.

I hope that this event was as much as valuable for the students that it was fot me and for the future of OTB.

Soon a “Coming next” article about recent improvements in the collaboration between OTB and Quantum GIS…

Simultaneously, we are releasing the new stable version of Monteverdi, the all-in one user-friendly graphical tool for remote sensing image processing and also a new release of OTB-Applications and OTB-Wrapping projects.

There are lots of exciting new features in this release!

These are the new main add-ons to the library:

• Add filter to perform arbitrary mathematical operation on image bands (BandMathImageFilter)
• Add class to parse mathematical expression (Parser)Add interpolation function (BCOInterpolateImageFunction)
• Add filter to compute the intensity of a complex image (ComplexToIntensityImageFilter)
• Add filter to compare two big images (StreamingCompareImageFilter)
• Add filters to compute extrema of big images (StreamingMinMaxImageFilter and StreamingMinMaxVectorImageFilter)
• Add filter to perform fine registration between images (FineRegistrationImageFilter)
• Add filter for higher order texture computation (ScalarImageToHigherOrderTexturesFilter and MaskedScalarImageToGreyLevelRunLengthMatrixGenerator)
• Removal of the old textures: http://wiki.orfeo-toolbox.org/index.php/Textures
• Add support for Worldview2 data
• Add margin sampler for SVM (SVMMarginSampler)
• Add generic resample filter that reproject in any arbitrary projection: cartographic, sensor model (GenericRSResampleImageFilter)
• Add class to produce a RPC model from a physical model (PhysicalToRPCSensorModelImageFilter)
• Add classes related to SAR calibration (SarParametricMapFunction, SarRadiometricCalibrationFunction, SarRadiometricCalibrationFunctor, SarRadiometricCalibrationToImageFilter)
• Add an action handler for the visualization enabling dragging (DragFullWindowActionHandler)

To the system:

• Internal ITK updated to 3.20.0 + OTB Patches
• Internal OSSIM and ossim plugins updated to svn revision 18162 + OTB patches (almost none left)
• Enable the use of an external build of libLAS
• Support compilation on MSVC 2010 and Windows Seven
• Better handling of FLTK configuration
• Support compilation of Monteverdi on linux 32 bits systems with official FLTK package
• Lots of coverage improvements

The Monteverdi application includes new modules:

• New Vectorization module: creates vector data (polygons, lines and points) from image
• New BandMath module: mathematical operation on image bands
• New SpectralViewer module: designed to display hyperspectral images. also computes spectral angle
• New ObjectLabeling module: object based image analysis module, with SVM classification based on object features
• New Fine Correlation module

The OTB-Applications package provide new useful command line tools:

• otbFastOrthoRectif: fast orthorectification based on grid subsampling
• otbActiveLearning: architecture to enable active learning application (put your own algorithms in the middle)
• otbCompareImages: image comparison (PSNR, RMSE…)
• otbFineRegistration: perform displacement estimation using the standard correlation metric between 2 images

A new release of the OTB-Wrapping project which allows to use OTB classes with other languages (Java and Python for now).

The complete list of new main add-ons is available here.

For MS-Windows users (XP/Vista/Seven) there are binary auto-install packages for both Monteverdi and OTB-Applications.

The OTB Software Guide has also been updated and is now 702 pages long.

As usual, you can download the source code and the binary packages from Source Forge. Follow this link for OTB and this link for Monteverdi.

Finally, we announce the opening of a new public English-speaking mailing list intended for OTB developers. All discussions related to the development activities, dashboard housekeeping, features to integrate and software architecture will now be held there. Please feel free to subscribe to the otb-developers mailing list and get involve into developing the Orfeo ToolBox!

The new OTB live DVD is out! On the program is the brand new OTB 3.4 (which came out 2 weeks ago) with the last Xubuntu 10.4.

It’s just as usual: download it from http://www.orfeo-toolbox.org/otblive/, burn it on a DVD, put it on a computer and restart (or launch directly  the iso file in a virtual host application like VirtualBox for example). A DVD is now required to contain the OTB library and all the associated projects (Monteverdi and OTB-Wrapping).

This is the screen you should get:

To help you get started with OTB, follow the guide on the help page:

You can also try the Monteverdi application or numerous OTB-applications directly out of the box:

You can also use the OTB-Python binding which are available through the OTB-Wrapping project. Just open a terminal and run the command python.

You can find detailed examples of OTB Python source code in the Software Guide or have a look to a previous post which put the emphasize on the powerfulness of the combination of OTB Python bindings and other open source Python project like PyLab. All these examples are ready to use in the last OTB Live DVD.

We hope that the new version is adapted as much as possible and will suit your requirements. Don’t hesitate if you’ve got suggestions or contributions about the project.

Give it a try!

These are the new main add-ons to the library:

• Produce hill shade image from DEM (HillShadingFilter)

• Compute point set density function (gaussian, epanechnikov)

• Extract a subset from a point set (PointSetExtractROI)

• Apply a transform to a point set (TransformPointSetFilter)

• Generate a random point set (RandomPointSetSource)

• Compute confusion matrix (ConfusionMatrixCalculator)

• Generate a list of samples from a vector data (ListSampleGenerator)

• Compute the ground spacing (GroundSpacingImageFunction)

• Read 6S spectral sensitivity files (SpectralSensitivityReader)

• Improvement in computation of Haralick Texture coefficients (ScalarImageToTexturesFilter, ScalarImageToAdvancedTexturesFilter)

Monteverdi application includes new  modules for region of Interest extraction, export image to Google Earth and a new module to perform supervised classification (Experimental).

Monteverdi can be considered ad the successor of the OTB-Applications package, but since some of the legacy applications are not yet integrated in Monteverdi, we are also shipping a 3.4 version of the applications. Moreover OTB-Applications provide also useful command line tools (otbConvert, otbConcatenateImage, otbSplitImage,otbExtractROI, otbConvertSensorToGeoPoint…)

For MS-Windows users (XP/Vista/Seven) there are binary auto-install packages for both Monteverdi and OTB-Applications.

The OTB Software Guide has also been updated and is now 702 pages long.

As usual, you can download the source code and the binary packages from Source Forge. Follow this link for OTB and this link for Monteverdi.

Don’t forget for those who will be attending IGARSS 2010 (July 25-30 in Honolulu) may be interested in the following tutorial:

“HD-2: Pragmatic Remote Sensing: A Hands-on Approach to Processing”

It will be given on Sunday, July 25, 08:30 – 12:30 and will present in detail the different steps of a general remote sensing image processing chain. Monteverdi will be used for the work.

As a consequence, the OTB Dev Team provides now materials which will demonstrate the software features in an interactive and didactic way.

Firstly, we now make available the slides used during  technical courses which give a complete overview of Monteverdi throught a tutorial which details available components.

Furthermore, the OTB Dev Team provides also screencasts which detail the way to construct interactive OTB process in the application.

Monteverdi provides a complete and operational environment to build interactively your own processes. This application takes transparently advantage of streaming and multi threading capacibilities of the Orfeo Toolbox pipeline. These tutorials will guide users to get a quick and operationnal description of the OTB/Monteverdi architecture and demonstrations of functionalities available for now.

These are first version of screencasts, lot’s of thing could be improve (for example, for now only the french version of the tutorials are available). Do not hesitate to give us feedbacks about them.

You can find direct link to OTB/Monteverdi slides.

You can find more details about the Monteverdi application and direct links to the Monteverdi screencasts.

And stay tuned on the Orfeo Toolbox Channel!

The number of users is growing very regularly (and that’s good!), the activity is not so regular (the drop for the summer and Christmas holidays are clearly visible), but overall, the increase in activity is obvious.