[GRASS-dev] GSoC introduction Roberta Fagandini

Wed May 16 07:22:06 PDT 2018

Hi Roberta,

Here in Norway we can do some “worst case” testing for your algorithm. Plenty of snow and clouds in the mountains, and heavily rugged terrain, that presumably has quite some effect on cloud shadows…

Looking forward to following your project from the side!

Kind regards,
Stefan

From: grass-dev <grass-dev-bounces at lists.osgeo.org> On Behalf Of Roberta Fagandini
Sent: onsdag 16. mai 2018 16.01
To: Pierre Roudier <pierre.roudier at gmail.com>
Cc: grass-dev <grass-dev at lists.osgeo.org>
Subject: Re: [GRASS-dev] GSoC introduction Roberta Fagandini

Thank you, Pierre!! I will keep the community constantly updated on the progress of the module.
Every feedback is welcome so please do not hesitate to send me yours! ;)

Roberta

2018-05-15 23:46 GMT+02:00 Pierre Roudier <pierre.roudier at gmail.com<mailto:pierre.roudier at gmail.com>>:
Interesting to hear your results, Roberta -- the reason I brought this
up is that some of my colleagues (non-GRASS users :( ) tried it very
successfully.

Happy to follow up with them if need be,

P

On 15 May 2018 at 22:03, Roberta Fagandini <robifagandini at gmail.com<mailto:robifagandini at gmail.com>> wrote:
> Hi Pierre!
> Thank you so much for your hints!
> I have already tested Fmask with Sentinel 2 images but I didn't have great
> results. However, it is worth investigating better!
> Thanks for all the references!
>
> Roberta
>
>
> 2018-05-15 0:51 GMT+02:00 Pierre Roudier <pierre.roudier at gmail.com<mailto:pierre.roudier at gmail.com>>:
>>
>> Hi Roberta,
>>
>> On top of the review linked by Vero, I thought I'd mention the Fmask
>> procedure -- it seems to give great results and there is a python
>> library on Github.
>>
>> *Relevant GRASS GIS tickets*:
>>
>> https://trac.osgeo.org/grass/ticket/3473
>> https://trac.osgeo.org/grass/ticket/3283
>>
>> *Papers*:
>>
>>
>> https://www.researchgate.net/publication/270596187_Improvement_and_expansion_of_the_Fmask_algorithm_Cloud_cloud_shadow_and_snow_detection_for_Landsats_4-7_8_and_Sentinel_2_images
>>
>> https://www.researchgate.net/publication/324836341_Improvement_of_the_Fmask_algorithm_for_Sentinel-2_images_Separating_clouds_from_bright_surfaces_based_on_parallax_effects
>>
>> *Software*:
>>
>> http://pythonfmask.org/en/latest/
>> https://github.com/prs021/fmask
>>
>> Hopefully this is helpful,
>>
>> Pierre
>>
>> On 7 May 2018 at 19:49, Roberta Fagandini <robifagandini at gmail.com<mailto:robifagandini at gmail.com>> wrote:
>> >
>> >
>> > 2018-05-06 21:52 GMT+02:00 Veronica Andreo <veroandreo at gmail.com<mailto:veroandreo at gmail.com>>:
>> >>
>> >> Hey Robi,
>> >
>> >
>> > Hi Vero!!
>> >
>> >>
>> >>
>> >> I just found this review [0]. It is for Landsat, but maybe some
>> >> insights
>> >> could be also useful for you (?)
>> >
>> >
>> > Thank you so much! I know this paper and it could be very useful
>> > especially
>> > for the second part of the procedure.
>> > I'll read it carefully!
>> >
>> >>
>> >>
>> >> Cheers :),
>> >> Vero
>> >
>> >
>> > Thanks!
>> > Robi
>> >
>> >>
>> >>
>> >> [0]
>> >>
>> >> https://www.researchgate.net/publication/324975294_Cloud_and_Cloud_Shadow_Detection_for_Landsat_Images_The_Fundamental_Basis_for_Analyzing_Landsat_Time_Series
>> >>
>> >> El jue., 3 may. 2018 a las 22:06, Roberto Marzocchi
>> >> (<roberto.marzocchi at gmail.com<mailto:roberto.marzocchi at gmail.com>>) escribió:
>> >>>
>> >>> Nice! The last step of the script you have written in python works as
>> >>> you
>> >>> expected.
>> >>>
>> >>> Now it is important to draw a diagram (or schema ) as a summary for
>> >>> you
>> >>> (you have worked a lot in the last few months) and to share it with
>> >>> Moritz
>> >>> and Markus.
>> >>>
>> >>> After that, test, test and test ;-) for validation/calibration of the
>> >>> automatic procedure.
>> >>>
>> >>> R
>> >>>
>> >>> 2018-05-03 18:48 GMT+02:00 Roberta Fagandini
>> >>> <robifagandini at gmail.com<mailto:robifagandini at gmail.com>>:
>> >>>>
>> >>>>
>> >>>> 2018-05-03 14:03 GMT+02:00 Moritz Lennert
>> >>>> <mlennert at club.worldonline.be<mailto:mlennert at club.worldonline.be>>:
>> >>>>>
>> >>>>> Hi Roberta,
>> >>>>
>> >>>>
>> >>>> Hi Moritz and Roberto!
>> >>>>
>> >>>>>
>> >>>>>
>> >>>>> On 25/04/18 18:03, Roberta Fagandini wrote:
>> >>>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>> 2018-04-25 16:03 GMT+02:00 Moritz Lennert
>> >>>>>> <mlennert at club.worldonline.be<mailto:mlennert at club.worldonline.be>
>> >>>>>> <mailto:mlennert at club.worldonline.be<mailto:mlennert at club.worldonline.be>>>:
>> >>>>>>     Looking at your bash scripts, I think the first thing to do
>> >>>>>> during
>> >>>>>>     this bonding period is, as you planned yourself, to get
>> >>>>>> familiar
>> >>>>>>     with the writing of GRASS modules in Python. You can have a
>> >>>>>> look
>> >>>>>> at
>> >>>>>>     existing scripts [1, 2] to get feeling for this works and how
>> >>>>>> to
>> >>>>>>     structure addon code in order to make it directly installable
>> >>>>>> with
>> >>>>>>     g.extension.
>> >>>>>>
>> >>>>>>     You can find the actual function definitions and documentation
>> >>>>>> of
>> >>>>>>     the GRASS Python scripting library at [3]. The functions in
>> >>>>>> that
>> >>>>>>     library should be more than enough to translate your scripts
>> >>>>>> into
>> >>>>>> a
>> >>>>>>     (or several) modules.
>> >>>>>>
>> >>>>>>     Be aware that GRASS modules create their own GUI. So, unless
>> >>>>>> you
>> >>>>>>     need some interactive features in your modules, you will not
>> >>>>>> have
>> >>>>>> to
>> >>>>>>     program your own GUI.
>> >>>>>>
>> >>>>>>
>> >>>>>> Thank you for your precious suggestions! I'll start studying how to
>> >>>>>> write a GRASS module in Python in the next days and at the same
>> >>>>>> time I will
>> >>>>>> keep on testing the procedures so as to show you some results and
>> >>>>>> fix some
>> >>>>>> open points.
>> >>>>>>
>> >>>>>>
>> >>>>>>     Something else you should probably do during this bonding time
>> >>>>>> is
>> >>>>>> to
>> >>>>>>     elaborate a schema of your algorithm, so that it is easier to
>> >>>>>>     understand what it does at each step.
>> >>>>>>
>> >>>>>>
>> >>>>>> Yes, this could be very useful also for me in order to better
>> >>>>>> organize
>> >>>>>> and put in order everything!
>> >>>>>>
>> >>>>>
>> >>>>> Have you advanced on any of this ? Do you have any questions ?
>> >>>>> Please
>> >>>>> don't hesitate to ask on the mailing list.
>> >>>>
>> >>>>
>> >>>> Yes, I started working with GRASS Python scripting library. I'm
>> >>>> following the link [0] you suggested, I'm also looking at other
>> >>>> existing
>> >>>> GRASS scripts [1,2] and moreover, Roberto gave me one of his scripts
>> >>>> as an
>> >>>> example. I have just committed the first version of the python script
>> >>>> I'm
>> >>>> working on, it works and I'm quite satisfied ;-)
>> >>>> Tomorrow I want to elaborate the schema of the algorithm and at the
>> >>>> same
>> >>>> time, I have to keep testing the procedure. As I wrote in the bash
>> >>>> file,
>> >>>> shadows detection seems to be strongly land cover dependent therefore
>> >>>> I
>> >>>> think it is necessary to test the procedure using several images
>> >>>> sensed in
>> >>>> different seasons, latitude, etc.
>> >>>>
>> >>>> Anyway, I'll commit some results tomorrow so as to show you something
>> >>>> more concrete!
>> >>>>>
>> >>>>>
>> >>>>> Best wishes,
>> >>>>> Moritz
>> >>>>
>> >>>>
>> >>>> Best regards,
>> >>>> Roberta
>> >>>>
>> >>>> [0]
>> >>>> https://grass.osgeo.org/grass75/manuals/libpython/script_intro.html
>> >>>> [1] https://trac.osgeo.org/grass/browser/grass/trunk/scripts
>> >>>> [2] https://trac.osgeo.org/grass/browser/grass-addons/grass7
>> >>>
>> >>>
>> >>> _______________________________________________
>> >>> grass-dev mailing list
>> >>> grass-dev at lists.osgeo.org<mailto:grass-dev at lists.osgeo.org>
>> >>> https://lists.osgeo.org/mailman/listinfo/grass-dev
>> >
>> >
>> >
>> > _______________________________________________
>> > grass-dev mailing list
>> > grass-dev at lists.osgeo.org<mailto:grass-dev at lists.osgeo.org>
>> > https://lists.osgeo.org/mailman/listinfo/grass-dev
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-dev/attachments/20180516/219c1a30/attachment-0001.html>