My Google Map Blog

Image recognition and Google Earth

by Timothy Whitehead on Jul.28, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

The last few years have seen major advances in computer artificial intelligence (AI). One area where AI is starting to show practical use is in imagery recognition. Google Earth and Street View imagery combined with image recognition has a wide range of possible applications. We have in the past had a look at Terrapattern, an experimental search engine for aerial and satellite imagery. They are adding new areas with time, so be sure to keep an eye on them.

We recently came across this story about a Caltech researcher that is helping the city of Los Angeles to count its trees with the help of a combination of Google Earth imagery and Street View. In this case they are trying to not only count individual trees but also identify the species.

The idea of using imagery for surveys of vegetation is of course far from new. Google Earth Engine, for example, is designed around such large scale analysis. When you wish to simply determine whether there is vegetation cover or possibly the overall health of the vegetation, a much better option than Google Earth imagery is to use false colour imagery – and satellites are typically designed with this in mind.

Another example of people using image recognition on Street View imagery is this one about identifying fire-hydrants and mentioned in that article is a project using Street View to study gentrification, which uses historical Street View to measure changes in buildings over time.

There is also this project, which uses Street View to geolocate an image. You could potentially take a photo with your mobile phone camera and the system could tell you where you were with accuracy similar to GPS. At present, this sort of thing is often done by crowd-sourcing rather than an automated system. The potential for automated systems has both potential benefits and serious privacy concerns.

Google itself applies some image recognition to Street View. The best known is identifying licence plates and faces, which are blurred for privacy reasons. However, it also reads house numbers and various street signs, and this information is used to improve Google Maps.


If Google were to add infrared to their Street View cameras, maybe it would make it easier to distinguish between faces of people who need privacy and faces of statues who need publicity.

Having infrared Street View has other uses and has been thought of already.

The post Image recognition and Google Earth appeared first on Google Earth Blog.

Comments Off :, , , more...

Baby sheep

by hragonezi on Jul.28, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off :, more...

Another way to visualise sun-synchronous orbits

by Timothy Whitehead on Jul.27, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Yesterday we had a look at the orbits of imaging satellite’s from the perspective of a stationary earth. Today we are having a look at the same orbits but showing how the orbit is actually a circle with the earth rotating inside it.

We found a model of Landsat 7 on the sketchup 3D warehouse and have created a tour showing what Landsat 7’s orbit looks like. The satellite is not shown to scale but the orbit should be approximately correct. Landsat 7 crosses the equator from north to south at about 10:00 am every 98.83 minutes (yes, it’s confusing). Its orbit covers the entire earth every 16 days and then repeats.

One problem we have encountered is animating a model across the antimeridian does not work correctly in Google Earth. We have not yet found a work-around. You will notice the model appears to jump occasionally when crossing the antimeridian. Another bug is that the background of stars shakes around when playing the tour. The stars should be stationary relative to the view, as the satellite’s orbit is nearly stable with respect to the stars, drifting approximately 1 degree per day (360 degrees per year).

Here we see Landsat 7’s orbit over the course of 24 hours:

You can view it in Google Earth with this KML file. For best results turn on sunlight (the icon with a rising sun on the toolbar). We also include in the KML the orbit for 24 hours or for the full 16 days.

Note that Landsat 8 shares the same orbit but with an 8 day offset.

This is what its 16 day orbit looks like relative to the earth:

We couldn’t record the full 16 day orbit as a tour as Google Earth couldn’t handle it. We believe it is possible to use a KML feature called a Track to improve performance, but we have not yet figured out how to do that.

For comparison, here is the layout of imagery tiles that are captured by Landsat 7 as provided by the USGS:

The post Another way to visualise sun-synchronous orbits appeared first on Google Earth Blog.

Comments Off :, , more...

Slow!

by sashafromdonetsk on Jul.27, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off :, more...

Sun-synchronous orbits with Google Earth

by Timothy Whitehead on Jul.26, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

After our recent posts on rainbow plane offsets and the list of imaging satellites we thought it would be interesting to see what a sun-synchronous orbit actually looks like in Google Earth. We have previously written a post about sun-synchronous orbits and why most imaging satellites use them, but we only showed an approximate single orbit of the earth assuming the earth was not rotating. In reality, the earth rotates on its axis as well as going around the sun. A sun synchronous orbit is designed to drift slowly so as to keep in sync with the earth’s orbit around the sun. All this starts to get complicated when you want to plot an actual orbit. But we believe we have succeeded.


What the orbit of WorldView-3 looks like.

We used the equations from Wikipedia, which allow us to use the altitude and period of a given satellite to work out its orbit. For simplicity we start at latitude zero and longitude zero. We show the orbit for a period of approximately 24 hours. Some satellites, such as Landsat 8, have their orbits arranged so that they repeat the same path on a regular basis. Others do not.

All sun-synchronous orbits look very similar, with the differences in altitude being hardly noticeable. The most obvious difference is in the period, which affects the spacing of the orbits.


Red: Landsat 8 orbit. White: WorldView-3 orbit.

WorldView-3 has an altitude of 617 km and a period of 97 minutes.
Landsat 8 has an altitude of 703 km and a period of 98.8 minutes.

The diagonals from the north east to south west ( / ) are on the daylight side of the earth and the diagonals from south east to north west ( \ ) are on the night side.

To see the above orbits in Google Earth download this KML file. Alternatively, you can create your own orbit by entering the altitude and period below:


Sun-synchronous orbit creator.

Altitude: km

Period: minutes

The post Sun-synchronous orbits with Google Earth appeared first on Google Earth Blog.

Comments Off :, more...

Topper for giants

by Willi1 on Jul.26, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off :, , more...

Imaging satellite list

by Timothy Whitehead on Jul.25, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Last week we mentioned that it would be useful to have a table of imaging satellites with some of the specifications of most interest to users of Google Earth. We couldn’t find any site with a single list in table form, so we decided to create our own.

The information comes from three main sources: Wikipedia, DigitalGlobe’s website and Sat Imaging Corporation’s website. We make no guarantees about the accuracy of this data.

Satellite Company Panchromatic
Resolution
Multispectral
Resolution
Launch
Date
Decommissioned Altitude
GeoEye-1 DigitalGlobe 46cm 1.64m 2008-09-06 681 / 770
WorldView-1 DigitalGlobe 50cm 2007-09-18 496
WorldView-2 DigitalGlobe 46cm 1.85m 2009-10-08 770
WorldView-3 DigitalGlobe 31cm 1.24m 2014-08-13 614
WorldView-4 DigitalGlobe 31cm 1.24m 2016-09 617
IKONOS DigitalGlobe 80cm 3.2m 1999-09-24 2015-04 681
QuickBird DigitalGlobe 55cm 2.16m 2001-10-18 2014-12-17 400
Pleiades-1A CNES 50cm 2m 2011-12-17 695
Pleiades-1B CNES 50cm 2m 2012-12-02 695
KOMPSAT-3A KARI 55cm 2.2m 2015-03-25 543
KOMPSAT-3 KARI 70cm 2.8m 2012-05-17 533
Gaofen-2 CAST 80cm 3.2m 2014-08-19 631
TripleSat (3 satellites) SSTL 80cm 3.2m 2015-07-10
SkySat-1 Terra Bella 90cm 2.0m 2013-11-21 450
SkySat-2 Terra Bella 90cm 2.0m 2014-07-08 450
SPOT-1 CNES 10m 20m 1986-02-22 1990-12-31 832
SPOT-2 CNES 10m 20m 1990-01-22 2009-07-01 832
SPOT-3 CNES 10m 20m 1993-09-26 1997-11-14 832
SPOT-4 CNES 10m 20m 1998-03-24 2013-07-01 830
SPOT-5 CNES 2.5m 10m 2002-05-04 2015-03-31 832
SPOT-6 CNES 1.5m 6.0m 2012-09-09 832
SPOT-7 CNES 1.5m 6.0m 2014-06-30 832
Landsat 1 USGS 1972-07-23 1978-01-06
Landsat 2 USGS 1975-01-22 1982-02-25
Landsat 3 USGS 1978-03-05 1983-03-31
Landsat 4 USGS 1982-07-16 1993-12-14
Landsat 5 USGS 1984-03-01 2013-06-05
Landsat 6 USGS 1993-10-05 1993-10-05
Landsat 7 USGS 15m 30m 1999-04-15 702
Landsat 8 USGS 15m 30m 2013-02-11 702
RapidEye (5 satellites) Planet Labs 5m 630
Doves (many satellites) Planet Labs approx 400
ASTER Japan METI 15m 15m 1999-12-18 705
Sentinel 2A ESA 10m 10m 2016-06-23 768
Sentinel 2B ESA 10m 10m 2016 768

Red: No longer active.
Blue: Not yet launched.

Notes on the above data:

Most of the satellites capture imagery in high resolution in grayscale (panchromatic) and lower resolution in each of the three primary colours and the images are combined to produce a high resolution colour image.

Orbits are typically not perfectly round and the earth itself is not perfectly spherical, so altitudes listed are averages. GeoEye-1 apparently had its orbit changed to a higher orbit in 2013.

Details not listed.

As far as we know, all the satellites can capture imagery in various infra-red bands that are typically lower resolution than the colour bands. There is significant variation between satellites as to which infrared bands they can capture.

There are differences in how the satellites capture imagery, such as whether they use a single sensor, a row of sensors or a sensor array.

Most satellites capture imagery in strips, which have a specific ‘swath width’ for a given satellite.

The best quality image is usually captured straight down from the satellite (nadir) but most satellites are capable of rotating so as to be able to capture imagery to either side of their orbit. How fast they can rotate varies by satellite.

Most satellites have a limit on how much imagery they can capture at a time. This is due to data storage and communication limitations.

Satellites vary as to how accurately they can report exactly where they were and what direction they were pointing when they took the photo.

The mass of the satellites varies considerably. In some cases the ‘satellite’ listed is actually an instrument on a larger satellite (ASTER for example).

Each satellite crosses the equator at a specific time.

Satellite design.

The various different requirements for satellites are
1. Resolution.
2. Acquisition speed. (How quickly a satellite can image a given area when required such as after a natural disaster).
3. Repeat acquisition. (The frequency with which it can image a given location).
4. Area covered. (Higher resolution typically means smaller coverage footprints, so lower resolution can be an advantage at times).
5. Consistency. (This is important for long term environmental monitoring).
6. Global coverage. (The ability to image the whole globe on a regular basis).
7. Price.

Commercial satellites tend to be more interested in 1. to 4., whereas government programs are more interested in 5. and 6. DigitalGlobe’s satellites are mostly expensive high resolution satellites whereas Planet Labs has sacrificed resolution in order to be able to make large numbers of cheap satellites, which allows them to cover more area quicker and repeatedly. The Landsat, SPOT and Sentinel satellites are designed for long term, regular, consistent coverage for environmental monitoring. The Sentinel satellites are essentially a continuation of the SPOT satellites with their orbits chosen to match.

Where does Google Earth imagery come from?

The main suppliers of satellite imagery to Google Earth at the current point in time are Digital Globe and CNES/Astrium, so almost all satellite imagery comes from their satellites. The low resolution global mosaic is Landsat 7 / 8 imagery, which is also used to fill in the gaps where there is no higher resolution imagery. Imagery from the SPOT satellites is also used for filling in gaps and can be recognised from the copyright notice “CNES/Spot Image”.

South America has a lot of imagery dated 1970. We are not sure which satellite it comes from, but we guess it may be from one of the early Landsat satellites and the date is only approximate.

Landsat, ASTER and Sentinel imagery is free to the public and can be downloaded and viewed in Google Earth.

Google Earth includes aerial imagery (captured from aircraft not satellites) from a number of sources.


Landsat 8. Image from Wikimedia commons.

The post Imaging satellite list appeared first on Google Earth Blog.

Comments Off : more...

Goats in Kyrgyzstan

by StreetViewFun.com on Jul.24, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off :, , more...

Anger issues

by StreetViewFun.com on Jul.23, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off :, , more...

Buy armor and become knights

by sashafromdonetsk on Jul.23, 2016, under 3D Models, Argentina, Australia, Brazil, California, Denmark, England, Germany, Google Earth News, Google Earth Tips, Google Sky, Google maps, Hawaii, Indonesia, Ireland, Italy, Japan, Kenya, Mexico, Natural Landmarks, Netherlands, Sightseeing, Street Views, USA

Comments Off : more...

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!

Visit our friends!

A few highly recommended friends...