[HOT] Mali Activation Follow-up

Harry Wood mail at harrywood.co.uk
Mon Feb 25 12:48:20 UTC 2013


To answer you question Stephane...

We have the GPX file coming from PovAddict's "Hunt the village" tool 

http://stuff.povaddict.com.ar/mali-crowdsource.gpx

(although not updated automatically)


The challenge is to remove from this set, any points which are villages/houses already in OpenStreetMap

We have places.csv  A set of locations of data already present in OpenStreetMap
http://harrywood.dev.openstreetmap.org/africaplaces/places.csv

I already succeeded in this challenge, by writing a brute force ruby script  http://harrywood.dev.openstreetmap.org/africaplaces/  Resulting GPX file:
http://harrywood.dev.openstreetmap.org/africaplaces/missingplaces_tm_198.gpx


...but it takes over an hour to run. Obviously this should have been done using PostGIS or GRASS or QGIS or basically anything other than my brute-force ruby script. So if you fancy that challenge, knock yourself out. However...

It's probably more important for people join in with these two ways of helping: http://bit.ly/XAJsuD   "Hunt the village" or "Map the village"!

...if lots of people join in with that, then it becomes more worthwhile to solve these other little things.


Harry



________________________________
 From: Stéphane Henriod <stephanebek at gmail.com>
To: Pierre Béland <pierzenh at yahoo.fr> 
Cc: HOT Openstreetmap <hot at openstreetmap.org> 
Sent: Monday, 25 February 2013, 3:33
Subject: Re: [HOT] Mali Activation Follow-up
 

Hi Pierre
>> People with some expertise in Spatial analysis in QGIS could be of great help to analyze data produced by the Crowdsource application. 
What kind of analysis are you looking for?
Cheers
Sent from Android
On Feb 25, 2013 4:35 AM, "Pierre Béland" <pierzenh at yahoo.fr> wrote:

Humanitarians are starting to access villages in north of Mali and we coordinate with them to support their activities.  HOT contributors are working on datafiles for various Public services, trying to geolocate these infrastructures.  We are working closely with OCHA coordinator in Bamako and started last week discussion with Clusters. The first discussion was with the WASH coordination Cluster for water and sanitation. We have exchange about our experiences geolocating various infrastructures and ways to improve coordinattion eand exchange of information.
>
>Tomorrow at 16:30 GMT a second discussion is scheduled with the Education Cluster on the irc   irc://irc.oftc.net/hot.
>
>
>In such a vast and complex territory to map, we have made substantial progress. But there is still a lot to do. Thanks to all of those who come on the irc, discuss and help in development of tools.
>
>
>Detecting villages in the vast and semi-desertic territory north of Mopti is still a priority. Nicolas Alvarez of HOT has proposed to use Crowsource of Imagery recognition for villages and dams and developped an application.  While this is still experimental, it gives very interesting results.  Pierre Giraud have adapted the Task Manager to accept Special micro-tasks defined from the result of this Crowdsource exercise. Instead of having contiguous micro-task tiles, we have tiles only where villages were detected.
>
>Others contributors are participate to the development on the  irc irc://irc.oftc.net/hot proposing various ways to deal with imagery recognition.  People with some expertise in Spatial analysis in QGIS could be of great help to analyze data produced by the Crowdsource application.
>
>Map that shows the result of the Crowsourcing at http://pgiraud.cartodb.com/tables/mali_crowdsource_waypoints/embed_map?title=true&description=true&search=false&shareable=false&cartodb_logo=true&sql=
>GPX file result of the Crowdsource tool at http://harrywood.dev.openstreetmap.org/africaplaces/missingplaces_tm_198.gpx
>
>You can contribute by mapping selecting tasks through the Tasking Manager. I invite you to test the
>Task Manager job based on these Crowdsourcing results.
>http://tasks.hotosm.org/job/198
>
>
>You can also pass the word that people with no experience can contribute by Crowdsourcing of Imagery recognition. This application is very simple. We either click on the image where we see a village, or click to move to the next image.
>See http://stuff.povaddict.com.ar/mali-crowdsource/
>
>
>
>Pierre 
>
>_______________________________________________
>HOT mailing list
>HOT at openstreetmap.org
>http://lists.openstreetmap.org/listinfo/hot
>
>
_______________________________________________
HOT mailing list
HOT at openstreetmap.org
http://lists.openstreetmap.org/listinfo/hot
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20130225/10b1a478/attachment.html>


More information about the HOT mailing list