I'm wondering what are acceptable sources for building heights, and especially how approximate we can be.

Some places have freely licensed data for building heights, derived from Lidar, or other sources. That's great, but when this is not available, what can we do?

I've just started to do some 3d mapping. To get height values, I've been estimating the ratio of height to width, using the measure tool in Josm to determine the width, and calculating a value for the height from these two. It's not a very accurate system.

Is it an acceptable approach, when lacking better quality data? Are there better systems?

asked 04 Jul, 18:29

keithonearth's gravatar image

keithonearth
2.4k425884
accept rate: 20%


I usually take pictures of the stuff I map, so I have used some guesstimates to determine heights.

I.e. I have a picture of a building im trying to map:

  1. It has a door, I assume the door is 2.2m high.
  2. Measure the height of the door in cm/pixels,
  3. Calculate how many real life meters a cm/pixel on the picture is.
  4. Measure the height of the building and muliply it by your value.

Now you have an estimate of the building height.

If you area has lidar-imagery available (and open) there is normally a point cloud where you can pan and zoom and rotate and look at objects in 3D and measure.

permanent link

answered 06 Jul, 09:22

FredrikLindseth's gravatar image

FredrikLindseth
661111829
accept rate: 13%

edited 06 Jul, 09:24

Personally, if I believe I can estimate it to the closest meter (which means it's pretty short or it's next to an object of known, similar height) I'll go ahead and add the integer height -- always with no decimal point, to indicate the lack of precision.

Failing that, I'll just add a building:levels=* tag. Whether that's any use to 3D mapping, I dunno, certainly the renderers could use it as a rough height estimate if they were so inclined. But if you're doing complex things like roof shapes, it's probably no help.

permanent link

answered 04 Jul, 21:59

jmapb's gravatar image

jmapb
75511218
accept rate: 16%

edited 04 Jul, 22:09

2

Mapping building:levels is very useful for 3D rendering, and it's indeed used by rendering software to estimate building heights. In areas where there are no legal sources for building heights, it's quite common to use building:levels almost everywhere and add heights only for the couple "special" buildings (towers, churches etc.) where that approach falls short. Of course, having heights for all buildings would be ideal, but counting levels is often a lot easier and therefore a good first step.

(05 Jul, 17:23) Tordanik

Good to know!

(05 Jul, 19:02) jmapb

There are free apps for phones measure the heights of buildings too. This is an example... https://itunes.apple.com/au/app/height-and-distance/id722926841?mt=8

permanent link

answered 06 Jul, 14:57

nevw's gravatar image

nevw
8.5k2483150
accept rate: 10%

edited 06 Jul, 15:15

Yeah I've tried that one, and a couple others. Comparing the results to tall buildings of known height, they are always wildly inaccurate. It's too bad because this is exactly the sort of magic that smartphones were hyped to be capable of. If there's an app out there that can do better, I'd love to try it. (I'm on iOS.)

(06 Jul, 16:53) jmapb

If you can't get an accurate measurement from an iPhone why not use old-school trignomentry? You can measure the distance to the base of the building (either physically, or calculate based on lat and long, and can calculate the height based on the measured distance and angle.

If you can get 5 degree accuracy with a cheap plastic protractor then you'll get the height of a 100m building at 100m away to within 20m, which is way more accurate than if you were sat on the roof with a regular GPS :)

permanent link

answered 06 Jul, 18:04

SomeoneElse's gravatar image

SomeoneElse ♦
30.5k62318707
accept rate: 15%

Yeah you can make a pretty good 3D model this way, if you have the patience. I would still doubt that you'd get sub-meter accuracy, so I'd advise to round to the nearest integer to avoid giving the impression that a height is an official or exact number.

Maybe consider tagging with source:height=calculated estimate, source:height=city building records, source:height=lidar, etc.

I've never had a GPS give me decent enough elevation that I'd trust it for building height. If I could get to the roof, I'd just drop a cannonball off the top and time how long it takes to hit the ground ;)

(06 Jul, 18:21) jmapb

I hadn't thought of using a plastic protractor, but I was imagining using a sextant to do this, and came to the conclusion that I really wasn't willing to go that far. It seems like there should be a way to get the angle from a photo, based on the focal length, and number of pixels from base to top of the building.

(06 Jul, 18:57) keithonearth
1

Using a walking pole say 1.2 metres measure poles shadow length then measure buildings shadow should give a very accurate estimation. If you have measured your pace length you can pace it out.

(07 Jul, 15:23) andy mackey
2

Building shadows on Bing can be used to estimate heights if an object of know height is in the same image.

(07 Jul, 15:26) andy mackey

This app as an angle read out that can be used with trigonmetry to get heights https://play.google.com/store/apps/details?id=com.digital_and_dreams.android.android_army_knife&hl=en_GB

(07 Jul, 15:39) andy mackey

Hi Keithonearth, why don’t you use the oldest method of measuring an object ? The pencil ? Look at the building alongside the pencil, mark the height and turn the pencil 90 degrees to the surface and mark the distance of the height and walk along it or measure it as well. The other method is a technical one, use a distance measurementtool with the option to calculate the height of an object using the method SomeoneElse mentioned (much cheaper). If so, choose an infra red one a simple one wont work that well during daylight or in the sun.

permanent link

answered 07 Jul, 09:30

Hendrikklaas's gravatar image

Hendrikklaas
8.4k170203340
accept rate: 6%

I estimated a tree height by measuring a nearby wall on a survey (level with third shirt button, for example). On bing the shadows of both can then be measured to calculate the tree height after seeing how high the button is.

(07 Jul, 15:33) andy mackey

Hi Andy, its the same method perfect & simple, but consider to take a spring rule with you its even usable as a large pencil.

(07 Jul, 22:23) Hendrikklaas

I have used a 3 metre tape but a 50 metre tape for tree shadows could cause some suspicion. I usually carry a backpack with GPS, camera, water, food, more clothes so i carry too much stuff already. But i agree old surveying and estimation methods are useful.

(08 Jul, 08:06) andy mackey
Your answer
toggle preview

Follow this question

By Email:

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

By RSS:

Answers

Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text](http://url.com/ "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:

×187
×27

question asked: 04 Jul, 18:29

question was seen: 518 times

last updated: 08 Jul, 08:06

powered by OSQA