In the database, the streets are expressed as a vector of nodes {points(lat, long)} ... So would you please explain the standards and methods which are depended on taking these specific points to forming the way? what about the distance between two points, and the number of nodes in each vector... Many thanks.

asked 23 Apr '18, 15:18

WijdanMTak's gravatar image

accept rate: 0%


It might help if you can give a bit of context to your question - what sort of answers were you expecting to get? What's the context in which you were asking it?

(19 Jun '18, 14:40) SomeoneElse ♦

There can be a maximum of 2000 nodes in a way.

That is about it.

permanent link

answered 24 Apr '18, 12:44

SimonPoole's gravatar image

SimonPoole ♦
accept rate: 19%

Thank you for your time, I really appreciate that.

That was helpful, but I want to know the minimum distance between two consecutive points. I would be grateful if you give me some details or references.

Best regards.

(19 Jun '18, 12:39) WijdanMTak

See Lat and lon both have 7 decimal places. Quoting from there: "The 7 rounded decimal places for coordinates in degrees define the worst error of longitude to a maximum of ±5.56595 millimeters on the Earth equator". So if I understand correctly the minimum distance is 5.5 millimeters and lower, depending on the specific longitude. There is a similar question at which has a slightly different answer: around 1 centimeter. According to 1 centimeter is correct for 7 decimal places.

(19 Jun '18, 12:47) scai ♦

There are no minimum/maximum standards for any element in OpenStreetMap. The data is sourced by millions of individuals with differing access to the information used to create the map (GPS tracks, visits,aerial and satellite imagery etc.) Therefore the choice of numbers of points will entirely depend on the preferences of the contributor. Some imported data may have many more points than is justified by the original accuracy of the source. So broadly speaking this is a meaningless metric.

For road geometries it is not possible to make any wide generalisation. However, in countries and areas which have been well-mapped with, typical several contributors to most road geometries, it is not unreasonable to anticipate the road centrelines will be within 5 m of the same feature where it is surveyed with more accurate methods and to a fixed standard by a national mapping agency. However, even in well mapped countries there will be some road geometries which have not been iteratively refined from the time when they were first added. Note the figure I give is not a proper statistical metric, but a rule of thumb.

There are numerous academic studies of road network quality of OSM going back around 10 years. These are probably the best place to: a) get a sense of the data quality; b) discover processes for your own assessment.

Certain types of roads are also likely to be less accurate than others:

  • Totally new roads, where no or limited aerial imagery exists. These may exist in the OSM data, but the actual precise geometry may have significant inaccuracies. Note it is usual that topological accuracy is maintained even if geometries are poor. For instance a new motorway may not being the right place, but all junctions will be with the correct intersecting roads. As such this data is usable for many purposes.
  • Roads in mountainous terrain where GPS signals and aerial imagery distortions mean that suitable sources for OSM are not adequate for deriving a precise geometry.
  • Rural areas of the USA with uncorrected TIGER data.
permanent link

answered 19 Jun '18, 14:31

SK53's gravatar image

SK53 ♦
accept rate: 20%

Your answer
toggle preview

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here



Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text]( "title")
  • image?![alt text](/path/img.jpg "title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Question tags:


question asked: 23 Apr '18, 15:18

question was seen: 489 times

last updated: 19 Jun '18, 14:40

powered by OSQA