Ebola Outbreak Nord-Kivu Building Geometry analysis

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Ebola Outbreak Nord-Kivu Building Geometry analysis

pierzen
I published a Blog post about a tool that Potentiel 3.0 develops with OSM-RDC to classify Building geometry and  to spot the more problematic contributions. Topological analysis is made from a PostgreSQL - PostGIS database.

I have looked at task 4947 that was completed august 6. I From the 18312 buildings in the area covered by this task, only 3.5% have irregular form (other then 90 degres or regular angles like huts).  There exist buildings with irregular form but they are the exceptions. 

The objective with the focus on the analysis of these irregular buildings, is one more angle to validate the OSM mapping. First, we can have rapidly a quantitative measure of irregular polygons to look at more closely. This also let's validate the more problematic contributions. 

The Overpass query links provide a list of buildings that are still reported as irregular today for the contributors with significant buildings flagged as irregular. Each query concern one contributor and a rapid look at each shows imprecise mapping. I would like the validators to look at this data and comment about how the focus on such data can help in the process of validation / correction of the mapping done for this Ebola response.

http://overpass-turbo.eu/s/AWc
http://overpass-turbo.eu/s/AWd
http://overpass-turbo.eu/s/AWe
http://overpass-turbo.eu/s/AWf
http://overpass-turbo.eu/s/AWg
http://overpass-turbo.eu/s/AWh
http://overpass-turbo.eu/s/AWi
http://overpass-turbo.eu/s/AWj
http://overpass-turbo.eu/s/AWk


regard
 
Pierre


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: Ebola Outbreak Nord-Kivu Building Geometry analysis

pierzen
If you are tired on the Talk list discussion about the gender of post codes, What's about OSM data quality ;)

Building Data quality Evaluation after Tasks are completed and validated is possible by the analysis of the Geometry of buildings.

For 3 jobs completed for the Ebola Response, you will find below links to Achavi. This will let you evaluate rapidly if this seems to satisfy OSM data quality standards.  Even if the Task manager indicates jobs as 100% validated, this seems to confirm that validators are demotivated and do not want to correct anymore imprecise building tracing from successive waves of one day contributors. 


https://overpass-api.de/achavi/?changeset=61276742
https://overpass-api.de/achavi/?changeset=61275198
https://overpass-api.de/achavi/?changeset=61348809


https://overpass-api.de/achavi/?changeset=61328699
https://overpass-api.de/achavi/?changeset=61326934
https://overpass-api.de/achavi/?changeset=61328228
https://overpass-api.de/achavi/?changeset=61329550




https://overpass-api.de/achavi/?changeset=61443529
https://overpass-api.de/achavi/?changeset=61444639
https://overpass-api.de/achavi/?changeset=61444931
https://overpass-api.de/achavi/?changeset=61442032
https://overpass-api.de/achavi/?changeset=61430587
https://overpass-api.de/achavi/?changeset=61430407
https://overpass-api.de/achavi/?changeset=61445212
https://overpass-api.de/achavi/?changeset=61444521
https://overpass-api.de/achavi/?changeset=26850026
https://overpass-api.de/achavi/?changeset=61424772
https://overpass-api.de/achavi/?changeset=61475962
https://overpass-api.de/achavi/?changeset=61430375
https://overpass-api.de/achavi/?changeset=61443320
https://overpass-api.de/achavi/?changeset=61475863
https://overpass-api.de/achavi/?changeset=61457626
https://overpass-api.de/achavi/?changeset=61457938
https://overpass-api.de/achavi/?changeset=61441962
https://overpass-api.de/achavi/?changeset=26836962
https://overpass-api.de/achavi/?changeset=61400757
https://overpass-api.de/achavi/?changeset=61445511
https://overpass-api.de/achavi/?changeset=61382566
https://overpass-api.de/achavi/?changeset=61459080
https://overpass-api.de/achavi/?changeset=61413766
https://overpass-api.de/achavi/?changeset=61459214
https://overpass-api.de/achavi/?changeset=61387946
https://overpass-api.de/achavi/?changeset=61458807
https://overpass-api.de/achavi/?changeset=61457301
https://overpass-api.de/achavi/?changeset=61478282
https://overpass-api.de/achavi/?changeset=61458243
https://overpass-api.de/achavi/?changeset=61438842
https://overpass-api.de/achavi/?changeset=61414264
https://overpass-api.de/achavi/?changeset=61445007
https://overpass-api.de/achavi/?changeset=61430147
https://overpass-api.de/achavi/?changeset=61384777
https://overpass-api.de/achavi/?changeset=61475989
https://overpass-api.de/achavi/?changeset=61421482
https://overpass-api.de/achavi/?changeset=61458523
https://overpass-api.de/achavi/?changeset=61441754
https://overpass-api.de/achavi/?changeset=61425864
https://overpass-api.de/achavi/?changeset=61389859
https://overpass-api.de/achavi/?changeset=26848029
https://overpass-api.de/achavi/?changeset=61443898
https://overpass-api.de/achavi/?changeset=61444362

regard
 
Pierre


Le jeudi 9 août 2018 16 h 14 min 25 s HAE, Pierre Béland <[hidden email]> a écrit :


I published a Blog post about a tool that Potentiel 3.0 develops with OSM-RDC to classify Building geometry and  to spot the more problematic contributions. Topological analysis is made from a PostgreSQL - PostGIS database.

I have looked at task 4947 that was completed august 6. I From the 18312 buildings in the area covered by this task, only 3.5% have irregular form (other then 90 degres or regular angles like huts).  There exist buildings with irregular form but they are the exceptions. 

The objective with the focus on the analysis of these irregular buildings, is one more angle to validate the OSM mapping. First, we can have rapidly a quantitative measure of irregular polygons to look at more closely. This also let's validate the more problematic contributions. 

The Overpass query links provide a list of buildings that are still reported as irregular today for the contributors with significant buildings flagged as irregular. Each query concern one contributor and a rapid look at each shows imprecise mapping. I would like the validators to look at this data and comment about how the focus on such data can help in the process of validation / correction of the mapping done for this Ebola response.

http://overpass-turbo.eu/s/AWc
http://overpass-turbo.eu/s/AWd
http://overpass-turbo.eu/s/AWe
http://overpass-turbo.eu/s/AWf
http://overpass-turbo.eu/s/AWg
http://overpass-turbo.eu/s/AWh
http://overpass-turbo.eu/s/AWi
http://overpass-turbo.eu/s/AWj
http://overpass-turbo.eu/s/AWk


regard
 
Pierre


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: Ebola Outbreak Nord-Kivu Building Geometry analysis

Ralph Aytoun-2
In reply to this post by pierzen

Hi Pierre,

 

One of the problems we face is that some very inexperienced mappers have been inadvertently marking task squares as validated and, while we are able to see in the Tasking Manager 3 how many squares that person has validated there is no mechanism in the Tasking Manager to identify those specific validated squares. This problem is being addressed at the moment with some urgency but, as with the Project 4947 that you used in your building analysis, although all the squares are validated some have not actually been checked properly and corrected.

 

Once the means is there to identify all the task squares in a project that have been validated by an individual we will be able to reduce the percentage of bad buildings even further.

 

I am glad to see that you are developing the means for us to identify the work done by an individual mapper as this will make it easier for validators to assess the standard of that mapper’s work overall and thus give them the appropriate helping hand and encouragement to improve their standard and become our more experienced mappers of the future.

 

 

Sent from Mail for Windows 10

 

From: [hidden email]
Sent: 11 August 2018 13:09
To: [hidden email]
Subject: HOT Digest, Vol 102, Issue 9

 

Send HOT mailing list submissions to

                [hidden email]

 

To subscribe or unsubscribe via the World Wide Web, visit

                https://lists.openstreetmap.org/listinfo/hot

or, via email, send a message with subject or body 'help' to

                [hidden email]

 

You can reach the person managing the list at

                [hidden email]

 

When replying, please edit your Subject line so it is more specific

than "Re: Contents of HOT digest..."

 

 

Today's Topics:

 

   1. weeklyOSM #420 2018-07-31-2018-08-06 (weeklyteam)

   2. Re: Ebola Outbreak Nord-Kivu Building Geometry analysis

      (Pierre Béland)

 

 

----------------------------------------------------------------------

 

Message: 1

Date: Fri, 10 Aug 2018 11:17:49 -0700 (PDT)

From: weeklyteam <[hidden email]>

To: [hidden email]

Subject: [HOT] weeklyOSM #420 2018-07-31-2018-08-06

Message-ID: <[hidden email]>

Content-Type: text/plain; charset="utf-8"

 

The weekly round-up of OSM news, issue # 420,

is now available online in English, giving as always a summary of all things happening in the openstreetmap world:

 

http://www.weeklyosm.eu/en/archives/10586/

 

Enjoy!

 

weeklyOSM?

who?: https://wiki.openstreetmap.org/wiki/WeeklyOSM#Available_Languages

where?: https://umap.openstreetmap.fr/en/map/weeklyosm-is-currently-produced-in_56718#2/8.6/108.3

 

------------------------------

 

Message: 2

Date: Fri, 10 Aug 2018 22:04:16 +0000 (UTC)

From: Pierre Béland <[hidden email]>

To: HOT Openstreetmap <[hidden email]>

Subject: Re: [HOT] Ebola Outbreak Nord-Kivu Building Geometry analysis

Message-ID: <[hidden email]>

Content-Type: text/plain; charset="utf-8"

 

If you are tired on the Talk list discussion about the gender of post codes, What's about OSM data quality ;)

Building Data quality Evaluation after Tasks are completed and validated is possible by the analysis of the Geometry of buildings.

For 3 jobs completed for the Ebola Response, you will find below links to Achavi. This will let you evaluate rapidly if this seems to satisfy OSM data quality standards.  Even if the Task manager indicates jobs as 100% validated, this seems to confirm that validators are demotivated and do not want to correct anymore imprecise building tracing from successive waves of one day contributors. 

Task 4947 https://tasks.hotosm.org/project/4947

https://overpass-api.de/achavi/?changeset=61276742https://overpass-api.de/achavi/?changeset=61275198https://overpass-api.de/achavi/?changeset=61348809

Task 4953 https://tasks.hotosm.org/project/4953

https://overpass-api.de/achavi/?changeset=61328699https://overpass-api.de/achavi/?changeset=61326934https://overpass-api.de/achavi/?changeset=61328228https://overpass-api.de/achavi/?changeset=61329550

Task 4958 https://tasks.hotosm.org/project/4958

 

 

https://overpass-api.de/achavi/?changeset=61443529

https://overpass-api.de/achavi/?changeset=61444639https://overpass-api.de/achavi/?changeset=61444931https://overpass-api.de/achavi/?changeset=61442032https://overpass-api.de/achavi/?changeset=61430587https://overpass-api.de/achavi/?changeset=61430407https://overpass-api.de/achavi/?changeset=61445212https://overpass-api.de/achavi/?changeset=61444521https://overpass-api.de/achavi/?changeset=26850026https://overpass-api.de/achavi/?changeset=61424772https://overpass-api.de/achavi/?changeset=61475962https://overpass-api.de/achavi/?changeset=61430375https://overpass-api.de/achavi/?changeset=61443320https://overpass-api.de/achavi/?changeset=61475863https://overpass-api.de/achavi/?changeset=61457626https://overpass-api.de/achavi/?changeset=61457938https://overpass-api.de/achavi/?changeset=61441962https://overpass-api.de/achavi/?changeset=26836962https://overpass-api.de/achavi/?changeset=61400757https://overpass-api.de/achavi/?changeset=61445511https://overpass-api.de/achavi/?changeset=61382566https://overpass-api.de/achavi/?changeset=61459080https://overpass-api.de/achavi/?changeset=61413766https://overpass-api.de/achavi/?changeset=61459214https://overpass-api.de/achavi/?changeset=61387946https://overpass-api.de/achavi/?changeset=61458807https://overpass-api.de/achavi/?changeset=61457301https://overpass-api.de/achavi/?changeset=61478282https://overpass-api.de/achavi/?changeset=61458243https://overpass-api.de/achavi/?changeset=61438842https://overpass-api.de/achavi/?changeset=61414264https://overpass-api.de/achavi/?changeset=61445007https://overpass-api.de/achavi/?changeset=61430147https://overpass-api.de/achavi/?changeset=61384777https://overpass-api.de/achavi/?changeset=61475989https://overpass-api.de/achavi/?changeset=61421482https://overpass-api.de/achavi/?changeset=61458523https://overpass-api.de/achavi/?changeset=61441754https://overpass-api.de/achavi/?changeset=61425864https://overpass-api.de/achavi/?changeset=61389859https://overpass-api.de/achavi/?changeset=26848029https://overpass-api.de/achavi/?changeset=61443898https://overpass-api.de/achavi/?changeset=61444362

regard 

Pierre

 

 

    Le jeudi 9 août 2018 16 h 14 min 25 s HAE, Pierre Béland <[hidden email]> a écrit : 

 

 I published a Blog post about a tool that Potentiel 3.0 develops with OSM-RDC to classify Building geometry and  to spot the more problematic contributions. Topological analysis is made from a PostgreSQL - PostGIS database.See https://opendatalabrdc.github.io/Blog/index.html#!Bulding_Geometry_Analysis_to_Support_OpenStreetMap_Quality_Analysis.md

I have looked at task 4947 that was completed august 6. I From the 18312 buildings in the area covered by this task, only 3.5% have irregular form (other then 90 degres or regular angles like huts).  There exist buildings with irregular form but they are the exceptions. 

The objective with the focus on the analysis of these irregular buildings, is one more angle to validate the OSM mapping. First, we can have rapidly a quantitative measure of irregular polygons to look at more closely. This also let's validate the more problematic contributions. 

The Overpass query links provide a list of buildings that are still reported as irregular today for the contributors with significant buildings flagged as irregular. Each query concern one contributor and a rapid look at each shows imprecise mapping. I would like the validators to look at this data and comment about how the focus on such data can help in the process of validation / correction of the mapping done for this Ebola response.

http://overpass-turbo.eu/s/AWchttp://overpass-turbo.eu/s/AWdhttp://overpass-turbo.eu/s/AWehttp://overpass-turbo.eu/s/AWfhttp://overpass-turbo.eu/s/AWghttp://overpass-turbo.eu/s/AWhhttp://overpass-turbo.eu/s/AWihttp://overpass-turbo.eu/s/AWjhttp://overpass-turbo.eu/s/AWk

 

regard 

Pierre

 

 

-------------- next part --------------

An HTML attachment was scrubbed...

URL: <http://lists.openstreetmap.org/pipermail/hot/attachments/20180810/4f99d903/attachment-0001.html>

 

------------------------------

 

Subject: Digest Footer

 

_______________________________________________

HOT mailing list

[hidden email]

https://lists.openstreetmap.org/listinfo/hot

 

 

------------------------------

 

End of HOT Digest, Vol 102, Issue 9

***********************************

 


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot