[SF Buildings] Mountain View

lxbarth commented 9 years ago

Prepare and lead a task to trace buildings in Mountain View, Los Altos and around. Some of it is already traced.

http://www.openstreetmap.org/#map=13/37.3828/-122.0870

Rub21 commented 9 years ago

@mapbox/team-data I've created task manager for this ticket. http://tasks.openstreetmap.us/project/9

@ediyes @luisalanya , ya pueden empezar con esta tarea :)

karitotp commented 9 years ago

Esta es la visualización de nuestro avance hasta el momento.

visualizacion mountain view

cc. @lxbarth, @Rub21

andreasviglakis commented 9 years ago

wow!

shvrm commented 9 years ago

amazing guys @mapbox/peru-office

planemad commented 9 years ago

:rocket: crazy stuff!

jothirnadh commented 9 years ago

The colour combination is so attractive... amazing work @mapbox/peru-office

ianshward commented 9 years ago

:eyes: Nice work!!! +1 on that color combo, it is spectacular.

On Saturday, November 21, 2015, Jothirnadh notifications@github.com wrote:

The colour combination is so attractive... amazing work @mapbox/peru-office https://github.com/orgs/mapbox/teams/peru-office

— Reply to this email directly or view it on GitHub https://github.com/mapbox/mapping/issues/136#issuecomment-158636292.

abhisheksaikia commented 9 years ago

This is incredible stuff @mapbox/peru-office

ruthmaben commented 8 years ago

Nice work @mapbox/peru-office!

oini commented 8 years ago

Wow! Great work @mapbox/peru-office.

bonniebogle commented 8 years ago

Great work everyone, this is fantastic!

ediyes commented 8 years ago

Done!!! :+1: buildingmv

cc @lxbarth @Rub21

lxbarth commented 8 years ago

:boom: !!

mikelmaron commented 8 years ago

As someone who has spent a lot of time there, I can say, Mountain View has never looked better!

On Tue, Dec 1, 2015 at 10:07 AM, Alex Barth notifications@github.com wrote:

[image: :boom:] !!

— Reply to this email directly or view it on GitHub https://github.com/mapbox/mapping/issues/136#issuecomment-160994856.

ramyaragupathy commented 8 years ago

Here is an overview of the review work for Mountain View buildings.

Team: @abhisheksaikia @nikhilprabhakar @jothirnadh @ramyaragupathy

Quality metrics

In Overpass, download buildings edited by the user in Moutain view. Open in JOSM.

[out:xml][timeout:25];

{{geocodeArea:Los Altos, USA}}->.searchArea;
// gather results
(
  node(user:"RichRico")["building"](area.searchArea);
  way(user:"RichRico")["building"](area.searchArea);
  relation(user:"RichRico")["building"](area.searchArea);
);

out meta;
>;
out meta qt;

In JOSM, click Search, add a new filter user:username building. Add selected buildings in todo list.
Visually inspect every building at >= z20.
Fix errors if needed.
Run JOSM validator for all selected buildings.
Upload edits.

user	# of buildings reviewed	correction visual check	correction via JOSM Validator
ediyes	4022	190	0
calfarome	1664	29	0
samely	1510	165	2
karitotp	2045	37	0
dannykath	98	15	3
RichRico	1419	10	0
Luis36995	77	3	0

Validating the completeness of the project

Also we reviewed the individual blocks in the TM to ensure the completeness of the project.

Observations

In some cases there is an inconsistency in using Mapbox/Bing as the source. Parts of the task uses Bing and the rest uses Mapbox interchangeably.
Sometimes an entire bunch of buildings is missing in one of the imagery but it is clearly visible in the other source.
Also in some cases the outline of the buildings are entirely different in the two imagery

In such cases the only option is to stick to the latest imagery. But the Mapbox imagery seems to be missing metadata on capture date, so we are not able to decide which image is the latest. For this validation process, we ran a history check of discrepant objects to decide the latest version.

Recommendations

For the SF peninsula building task, it would be great if the team just sticks to Bing imagery as the source (this seems to be the latest in most places) and fall back to Mapbox imagery only in case of non-availability of Bing imagery.

cc/ @geohacker @planemad @rub21 @lxbarth

lxbarth commented 8 years ago

Awesome review post.

In some cases there is an inconsistency in using Mapbox/Bing as the source. Parts of the task uses Bing and the rest uses Mapbox interchangeably.

@ramyaragupathy do you see a need to go back and review the accuracy of mapping in the cases where Mapbox imagery was used?

@ediyes - in which cases did the team use Mapbox versus Bing imagery?

abhisheksaikia commented 8 years ago

@lxbarth I think in this case we should wait for the satellite team to get back on the ticket we raised for metadata information in Mapbox imagery.

@ramyaragupathy do you see a need to go back and review the accuracy of mapping in the cases where Mapbox imagery was used?

At present there is no sure way to determine which among Bing and Mapbox has the more updated imagery as Mapbox is missing Metadata information. Until we have this information, we should hold on going back and reviewing the task.

ramyaragupathy commented 8 years ago

Until we have this information, we should hold on going back and reviewing the task.

:+1:

Also @lxbarth we don't have a log of places that used Mapbox imagery in the task.

ediyes commented 8 years ago

@ediyes - in which cases did the team use Mapbox versus Bing imagery?

@lxbarth todos utilizamos el Bing, pero habian edificios ya hechos en algunos lugares y para mejorarlo utilizamos la imagen de Mapbox cuando no existia el Bing.

mapbox / mapping