>>
you're reading...
Art-Science

Radar images as you’ve never seen them before…

Example of icon based visualisation of polarimetric data.

An example of how the icon-based visualisation allows you to explore the polarimetric information contained within the image, while maintaining the geographical context.

In an earlier post (What price, creativity?) I talked about the value of artistic creativity in innovation.  Here is a great example of this in action…

We were inspired by the fabulous work coming out of Brown University, using artist painting techniques to create new ways of visualising multidimensional data sets (the link to the full paper based on Dean Turner’s PhD is below).

The challenge we were exploring was that radar imagery (of the kind we frequently use to measure forest properties  –  that’s the forest link here for this blog!) records the polarisation of the echoes.  The polarisation is the orientation of the electromagnetic waves, and it gives you information on the orientation or physical structure of the object on the ground.  Its a very useful tool, but it does mean you can end up with at least five “pictures” that you want to visualise all at once – only three of which naturally equate to a “brightness”. The other parameters relate to orientation angles and asymmetries, which traditional image generation can’t handle too well.

A standard approach to making a colour image from remotely sensed data is to have three channels of brightness information – essentially three separate black and white images, each created using a different wavelength of light.  You assign each brightness level to the Red, Green or Blue channel of the display, and the various combinations of brightness of these different hues gives a whole range of resulting colours[1].So how do you best visualise an orientation angle? A brightness value is not a very good way to represent it – a brighter image does not intuitively match to a larger angle – and the fact that the highest angle (360°) actually equates to the lowest angle (0°) doesn’t match very well onto brightness (“the brightest pixel equals the darkest pixel” isn’t very intuitive).

Another example of the icon-based visualisation

Another example of the same technique, this time over San Francisco.

So what we have done is used little “icons” to represent various parameters of the data. At low resolution, you still see the broad geographical context, but if you look closely at a region of interest you see that it is made up of little ellipses that represent the polarimetric information.It looks beautiful, but it also communicates the polarimetric information in a more intuitive manner and allows the user to visually explore the data more effectively.


The full article is available on Open Access here:

Dean Turner and Iain H. Woodhouse
An Icon-Based Synoptic Visualization of Fully Polarimetric Radar Data
Remote Sens. 20124(3), 648-660; doi:10.3390/rs4030648
http://www.mdpi.com/2072-4292/4/3/648/

Other links:

David Laidlaw’s Visualization Lab at Brown University


[1] If you take a magnifying glass and look at a TV screen, you will see that each pixel is composed of a red, a green and a blue light.  If they are all lit equally brightly, it looks white. Other colours are created from various levels of brightness of each channel

Advertisements

Discussion

5 thoughts on “Radar images as you’ve never seen them before…

  1. Hello,

    Looks nice, but it is not explained what is the meaning of colors, ellipse sizes and the polarizations shown, i.e. are they transmit or receive polarizations or something else.

    Cheers,
    Tiitu.

    Posted by Tiitu | March 15, 2012, 4:40 pm
  2. Hi Tiitu,
    The link to the full article is at the bottom of the post. Here you will find the explanation for the ellipses size, shape and colour.
    http://www.mdpi.com/2072-4292/4/3/648/

    -iain-

    Posted by fortiain | March 15, 2012, 4:44 pm
  3. thanks for sharing. this is such a new technique!

    Posted by kedaimp3player | August 1, 2012, 10:06 pm

Trackbacks/Pingbacks

  1. Pingback: No reason not to publish Open Access « Forest Planet - April 11, 2012

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow me on Twitter

%d bloggers like this: