OpenStreetMap 로고 OpenStreetMap

nammala의 일기

최근 일기 항목

What's new in OSMCha

nammala님이 English로 2017년 10월 11일에 게시함. 최근 2017년 10월 16일에 업데이트됨.

If you have been a daily user of OSMCha, you have already started using the new version. Wille has been working with Mapbox on new developments of OSMCha. We have been testing the stack stability, and collecting feedback from Mapbox data team and community members for few months now.

I would like to use this diary post to introduce few features in the new version of OSMCha and talk about ways we can use them for validation.

Introduction

The development challenge has been to make OSMCha easier to use, faster to review changesets, and a robust tool for filtering OSM edits. We needed to add contextual information on mappers & changesets, improve the user-friendly UI, add keyboard shortcuts, and a lot more using an API driven frontend.

Along with a brand new design, a lot of things have been improved in the new version. Here are some of the new features.

Sidebar

모든 항목 보기

Validation Scoop: Inconsistent edits observed

nammala님이 English로 2017년 5월 26일에 게시함.

Resuming from previous diary posts, we would like to start off with interesting changesets we have observed over the last month. Nearly all of these changesets have been flagged by currently existing filters in OSMCha.

Here is what we found

image We found randomly added parks in the form of letters K N N in this changeset. This was detected while we were reviewing changesets from pokemon nest by new user filter in OSMCha.

모든 항목 보기

Continuing from our previous diary post on inconsistent edits observed, here are the inconsistent edits observed in OpenStreetMap between 26 December 2016 - 27 January 2017 using OSMCHA.

We commented on the changesets:

  • Deleted main stream and added an incomplete one: changeset
  • Deleted existing roads and buildings: Changeset 1, 2, 3
  • Added unnecessary trunk roads: Changeset
  • Deleted buildings: Changeset 1, 2, 3
  • Deleted roads: Changeset 1, 2
  • Added unconnected footways with footway=crossing, crossing=zebra for the whole length of the road: Changeset
  • Added self intersecting footway that goes over a building: Changeset
  • Added address in name tag: Changeset
  • Added single large buildings for many buildings instead: Changeset

Community members commented on the following changesets:

Reverted:

  • Added waterbody which doesn’t exist in the imagery: changeset
  • Added christmas tree saying Happy new year: changeset

모든 항목 보기

Following up from our previous diary post on inconsistent edits observed, here are the inconsistent edits observed in OpenStreetMap between 28 November - 23 December using OSMCHA.

We commented on these changesets:

  • Deleted road: changeset 1, 2, 3, 4, 5
  • Added inappropriate name tag to the highways: changeset
  • Added building=yes to nodes for some features: changeset
  • Added lots of POI’s: changeset
  • Deleted rail road: changeset
  • Buildings that were not split: changeset
  • Circular polygon over university: changeset

Community members commented on the following changesets:

Reverted:

  • Deleted buildings: Community commented and reverted the changesets 1, 2, 3, 4
  • Deleted roads: Reverted the changeset 1, 2, 3, 4, 5, 6, 7, 8
  • Overlapping buildings: Reverted the changeset
  • Added a ton of ways with relation type=lanelet at intersections. DWG reverted the changeset
  • Fantasy mapping. Community reverted the changeset 1, 2
  • Deleted landuse. Reverted the changeset
  • Deleted track: Community reverted the changeset
  • Mass incorrect edits of road graph. We reverted 52 changesets
  • Adding bad import hotel data. Community reverted the changeset
  • Added POI to a major road. Community reverted the changeset
  • Deleted tunnel on a significant highway. Reverted the changeset
  • Added private information. Reverted the changeset

If you are interested in validation, check out our validation guide on using OSMCha. Do keep an eye out and comment on changesets which seem inconsistent to let the user know which will in turn helps users become better mappers and lets us maintain data quality on OpenStreetMap.

Look forward for another roundup next week.

Here are the few observations from the OpenStreetMap edits between 11 November - 25 November. We looked into the filters like mass deletions, iD editor + mass deletions, possible imports, edited a name tag, mass modifications using OSMCHA for reviewing the changesets.

Commented:

  • Deleted tracks: changeset
  • Added random pedestrian highways and buildings: changeset
  • Deleted highways and few buildings: changeset 1, 2, 3
  • Deleted buildings and some amenities: changeset 1, 2, 3
  • Incorrect tagging: changeset
  • Changed road classification: changeset
  • Deleted neighborhood tags: changeset
  • Added Korean names in name:en tag: changeset
  • Deleted streams: changeset

Community members commented on the following changesets:

Reverted:

  • Added province tags to address: Community member commented and reverted the changeset
  • Undiscussed import: Community member commented and reverted the changeset 1, 2
  • Deleted houses and residential roads: Reverted the changesets with a comment 1, 2, 3
  • Deleted buildings and roads. Community member commented and reverted the changeset.
  • Added orphan nodes over buildings and highways. we reverted the changeset with a comment

These were some of the inconsistent data for this week. Do keep an eye out and comment on changesets, which will make us maintain the quality of data in OpenStreetMap.

Look forward for another roundup next week.

Here are the few observations from the OpenStreetMap edits between 24 October - 11 November.

Commented:

  • Bad imports: changeset 1, 2, 3, 4
  • Deleted roads: changeset
  • Deleted natural=wood: changeset
  • Deleted existing buildings: changeset 1, 2, 3, 4
  • Adding fictional data: changeset 1, 2
  • Adding buildings overlapping with highways: changeset
  • Changeset comment mentioning google: changeset
  • Deleted river: changeset
  • Added improper data: changeset
  • Deleted relevant information in neighborhood: changeset 1, 2

Community members commented on the following changesets:

These were some of the inconstant data for this week. Do keep an eye of the bad edits and comment on those changesets, which will make us maintain the quality of data in OpenStreetMap.

Look forward for another roundup next week.

One of the most important uses of the map during car navigation is to find the correct exit off a motorway to get to a destination. Missing an exit can be quite costly in terms of the time and fuel wasted in a much longer route than desired. In OSM, the tags that help to map this information are highway=motorway_junction(exit number) and destination(destination sign).

The Mapbox data team has updated over 247 motorway exits and 1285 destination sign information in United States & Canada in the last year. Here’s the workflow we followed to create tasks on the Mapbox Tasking manager if you would like to create a community project for your own area. This workflow creates convenient tasks that can be distributed to review the data on individual highway routes on checkautopista2 for a particular area of interest.

screenshot 2016-11-14 13 12 21

1. Defining the area of interest:

Draw a boundary for the area of interest you want to cover in JOSM tagging as area=yes and save it as .osm file.

모든 항목 보기

Imagery offset database in JOSM

nammala님이 English로 2016년 10월 24일에 게시함.

Satellite imagery is one of our common source of data for mapping, however, not all imagery is exactly aligned to real world features. Every mapper should be aware of this and should consider adjusting imagery from aggregated GPS tracks or from features which we know have correct location.

Adjusting imagery offset

The usual approach is to align the imagery to GPS traces as best as possible using the Adjust imagery offset tool before making any edit. screen shot 2016-10-20 at 11 53 08 am

모든 항목 보기

Continuing from our previous weekly round up, these are the observations between 12 September - 30 September.

Commented:

  • Deleted buildings: changesets 1, 2, 3.
  • Deleted admin boundary of region in Burma: changeset.
  • deleted mosque: changeset.

Community members commented on following changesets:

  • Adding numbers on the name tags for lot of buildings: changeset.
  • Renamed a department store in Italy to multilingual name tag with Chinese added: changeset.
  • Deleted name tag of a multipolygon and other features: changeset.
  • Deleted lake that exists in the imagery: changeset.

Reverted:

  • Deleted tertiary roads: changeset community member reverted the changeset.
  • Bad imports: changeset. We reverted the changeset.
  • Uploaded learnOSM test data to the map. changeset community member reverted the changeset.
  • Added fictional roads and deleted many existing highways & other features: changeset. We reverted the changesets.
  • Deleted buildings in Philippines. We reverted 4 of his changesets 1, 2, 3, 4.
  • Added natural=water to all the nodes of the lake. changeset. We deleted the edits with proper comment.

These were some of the inconsistent edits for this week. Do keep an eye and please comment on such changeset, this will help in maintaining the data on OSM accurate and coherent.

Look forward to another roundup next week.

Progress update of missing turn restrictions in Germany

nammala님이 English로 2016년 9월 19일에 게시함.

Last week, as a part of making OpenStreetMap more navigable, we started updating possible missing turn restrictions in 5 cities of Germany i.e., (Berlin, Stuttgart, Wolfsburg, Munich and Frankfurt), with the help of OSM Navigation map which uses Mapillary detected traffic signs. Around 1900 turn restriction were reviewed in these 5 cities and we managed to add 112 OSM notes where probable missing data was found.

screen shot 2016-09-19 at 3 00 19 pm

We have received a great response from the community on this task, and as of now, 71 notes are resolved. It would be great if the community could resolve remaining 41 notes to enhance the accuracy of navigation data present on OpenStreetMap.

Here is the list of OSM notes for missing turn restrictions.

모든 항목 보기

Updating Navigation data in Germany

nammala님이 English로 2016년 9월 9일에 게시함.

As a part of making OpenStreetMap more navigable and accurate for routing, we started updating missing turn restrictions and exit numbers & destinations in 5 cities in Germany: Berlin, Stuttgart, Wolfsburg, Munich and Frankfurt. We started off with 3 cities and later jumped into other 2 cities, Munich & Frankfurt and managed to update missing turn restrictions and exit numbers & destinations through OSM notes with the help of community support.

Summary of improvements:

The Mapbox data team started updating turn restrictions on August 26. The The turn restrictions signboard conventions in Germany were quite different from what we followed for Canada and United States. We went through the German wiki to understand the tagging system and carried out our initial research and came up with our observations in this diary post, where we received valuable feedback from the community. We came to a conclusion that adding OSM notes for the missing turn restrictions is a good way for the community mappers to verify the data with their local knowledge.

We used OSM Navigation map which uses Mapillary images to detect the missing turn restrictions for this. Around 1900 turn restrictions were reviewed in these 5 cities and 112 OSM notes were added by the team where possible missing data was found.

For the Exit numbers and destinations task, the team followed the wiki to review and added OSM notes for the missing exit numbers and destinations for the community to verify and map. For this process, we used tasking manager and checkautopista2 for identifying the missing exit numbers and destinations.

Progress update of missing turn restrictions:

모든 항목 보기

Germany navigation data review update

nammala님이 English로 2016년 9월 1일에 게시함.

Following the previous diary post on improving navigation data in Germany, around 1000 turn restrictions detected on Mapillary were reviewed and 72 notes were added on OSM where there was a mismatch with the data for community mappers to verify with local knowledge. The community response has been extremely quick to verify an resolve these notes almost within a day of them being added. The whole exercise resulted in adding 27 new turn restrictions to the map which is already highly detailed and up to date.

screen shot 2016-09-01 at 8 42 44 pm (Blue=added to OSM; Orange=not required to add to OSM; Red=Invalid detection or location)_

Exit and destination signs on motorways

모든 항목 보기

This diary post last week outlined an overarching goal of making OpenStreetMap navigable, and our intent to investigate and add missing turn-restrictions in Berlin, Stuttgart, and Wolfsburg using Mapillary as a source. We received some good questions from the local OpenStreetMap community about using Mapillary as a source. Considering the concerns of the community, we did a preliminary research on the Mapillary imagery to ascertain their recency. We want to share our findings here.
Here is an animation showing Mapillary traces in Berlin, Stuttgart, and Wolfsburg spanning the last six months (March - August 2016). Not surprisingly, Berlin was found to have the best Mapillary coverage. The coverage in Stuttgart and Wolfsburg was sparse in comparison.

Berlin2 Berlin Mapillary Coverage

모든 항목 보기

Weekly roundup of edits

nammala님이 English로 2016년 8월 19일에 게시함.

Here are a few observations from the OpenStreetMap edits between 8 August - 19 August. Most of these helped us understand common mapping issues faced by the new mappers.

  • Added untagged nodes 1, 2. These were reverted by the community.
  • Added nodes with numbers as names 1. Reverted by community.
  • Bad imports of data 1. Reverted by commuinty.
  • Deleted waterway=riverbank , natural=water tags and stream 1, 2, 3 respectively. Asked for clarification.
  • Deleted building 1. Asked for clarification.
  • Deleted highway=service roads inside the park 1. Asked for clarification.

It is always good to let the user know of such issues and encourage them to do active and quality contributions to OpenStreetMap.

Looking forward to another roundup next week. Happy Mapping.

Mapping missing turn restrictions & exit numbers and destinations in Europe - Part 1 : Germany

nammala님이 English로 2016년 8월 12일에 게시함. 최근 2016년 8월 19일에 업데이트됨.

Turn restrictions, exit and destinations are important components of navigation system and enhance the routing and guidance accuracy of any routing engine. To improve OpenStreetMap, broaden its reach globally and make it more accurate for routing, we would like to join hands with the community to map missing turn restrictions, exit and destinations in Europe using Mapillary as the primary source.

pasted image at 2016_08_19 04_49 pm

Timeline

First stop is Germany, we want to concentrate on three major cities: Berlin, Stuttgart, Wolfsburg and map turn restrictions, exit and destinations, slowly branching out to other cities and countries in Europe.

Mapping workflow

Turn Restrictions

We have devised a simple way to add turn restrictions using the Mapillary and detected traffic signs using OSM navigation map that displays the detected turn restrictions from Mapillary, making it easy to know the position of the signage and to mark it as valid/invalid once it has been reviewed and mapped if required.

모든 항목 보기

Understanding common mapping issues is helpful to know difficulties faced by new mappers and can give insight into improving the editors for a better mapping experience. Here are some of the issues observed between 4 July - 15 July on the map:

  • building=house tags added to the nodes of the buildings: 1, 2. These were reverted.
  • Roads without any tags: 1. Asked for clarification.
  • Deletion of highway=track ways without explanation or source: 1 . Asked for clarification.
  • Roads were deleted as they are ‘not public’: 1, 2. Community responded asking for clarification.
  • Duplicate buildings & incorrect tags: 1, 2, 3. Community responded with a comment.
  • Deleted roads: 1, 2. Commnunity responded with a comment.
  • Deleted landuse=residential tags: 1, 2. Asked for clarification.
  • Deleted highway=primary roads: 1. Asked for clarification.
  • Overlapping buildings: 1. Community responded with a comment.
  • Added duplicates of few buildings: 1. Was fixed by the community.

It is always good to let the user know of the mistake to encourage active, qualitative contributions to OpenStreetMap.

Look forward to another roundup next week.

Mapping Destinations Signs on US Motorways

nammala님이 English로 2016년 5월 5일에 게시함. 최근 2016년 5월 6일에 업데이트됨.

The data team at Mapbox has been mapping destinations signs in the US. Based on Mapillary images, Department Of Transportation (DOT) documents, crosscountryroads, and Interstate guides.

image

Imagery source- Mapillary

We started working on 9 states:

  • California
  • Florida
  • New Jersey
  • Maryland + DC
  • Pennsylvania
  • Washington state
  • Colorado
  • Texas
  • Illinois

We are using checkautopista for adding destination tags like destination,destination:street,destination:ref after verifying from the above sources.

The team worked on this task last month and reviewed 220 highways and added 3,417 tags

  • Number of destination tags added: 1663 (map)
  • Number of destination:ref tags added: 1481 (map)
  • Number of destination:street tags added: 273 (map)

This is the workflow we followed.

Happy Mapping!

Adding POIs and cleaning up data in Visakhapatnam

nammala님이 English로 2016년 3월 16일에 게시함.

Visakhapatnam is my hometown, a coastal city in Southern India, it is popular tourist spot with temples, hills, parks and beaches. This city is also home to Indian Navy’s Eastern Naval Command. A constantly evolving city, Visakhapatnam ranked 8th in the smart city list of the Government of India for futuristic infrastructure development.

vizag_aerial_view

Aerial photographs of Visakhapatnam by Candeo gauisus, CC BY-SA 3.0

I have done a lot of mapping of my hometown in the last 4 months. But after going through several mapping guides and discussion with my colleagues at Mapbox, I realized that I can improve my mapping even further.

So, I started cleaning up my earlier edits and added points of interest (POIs) with proper tagging and naming practices.

This is how I went about the whole process:

모든 항목 보기

위치: Indiranagar 1st Stage, Indiranagar, Bengaluru Central City Corporation, Bengaluru, Bangalore North, Bengaluru Urban, Karnataka, 560038, India

Mapping Andhra University College of Engineering (Andhra Pradesh, India) using Field Papers:

nammala님이 English로 2016년 1월 28일에 게시함. 최근 2016년 1월 30일에 업데이트됨.

Hello OSM mappers,

I ‘m new to OSM community and this is my first diary post in which I want to explain my mapping experience in using field papers

As a part of mapping using Field Papers for which I selected an area in my neighborhood (Visakhapatnam), I have chosen Andhra University college of engineering. The main reason for selecting this region as area of interest (AOI) is, I had done my bachelors from this University and also it is not yet mapped in OSM.

모든 항목 보기