Home > Photography > Panoramas


Manasquan Boardwalk 2Manasquan Boardwalk 1

Two Fisheyes

Place is all important, there are so many written documents about people, places and things that it is hard to get a handle on anything. Enter the GIS, that allows us to visualize what is happening around us into a spatial picture. If we want to learn about an area we can open Google Earth, Flickr, Panoramio or other mapping tool. Besides these public sites and tools most people and most organizations have a plethora of other images and experiences that are tied to locations. If these images and sounds are placed in an environment that allows the users to interact with them, a form of virtual reality is created. We can gain a situational awareness of an area without actually going to the location.

I have been playing around with how to leverage the power of place in regard to hundreds of pictures of different locales. Especially interesting is how can we create the “immersive hotdog” experience that is postulated on Greetings from Manasquan’s blog. More particular for me, how can I create this immersing experience with free / non server software, which can preferably be used without having to install it on the computer.

Panoramas have always been a fascinating to play with. In the past I have explored panoramas using AutoStitch, from Matthew Brown, University of British Columbia. Now I am looking at several pictures that were taken with a fisheye lens for use with the iPix software. However, I do not have a copy of the software but would still like to use the pictures.

AutoStitch does not have the capabilities to convert fisheye pictures to traditional rectangular pictures. I have been exploring several different options, however I have not yet figured out how to do this. During my research I came across a website which promises to defishify a picture, however the downloaded program is a .sit file. I believe this is the main compression type for Macs, I use Windows. The website does say that Windows 98 and NT are supported but I have had little success in finding a program that I can use to uncompress the files.

I have found, Panorama Tools, which seems like it can help. However I am still trying to understand how to use the software. Lately I have been playing around with hugin which is a manual panorama creator. It promises that it can take a fisheye picture and turn it into a rectangular picture. I can then stitch the pictures together. However my results have been less than spectacular. The software has issues with perspective, and I’m not sure the “immersive hotdog” experience is going to be possible using fisheye pictures.

Categories: Photography Tags: , ,
  1. geosquan
    January 1, 2008 at 6:41 pm

    This is great! Thanks for playing with the data and that like me you want this to be a free software solution.

    One thing of note the Input 1 and Input 2 you use above are not intended by IPIX to stitch together, even though they seem logical to do so.

    Remember that those are successive shots which each have corresponding “back side” images.

    That is how iPIX wants you to stitch them – as I have done with their software and exported to Equirectangular jpegs here


    But you’re right, what would be nice would be to be able to also stitch panoramas along a plane such as your Input 1 and Input 2 above.

    As I moved down the boardwalk and changed my center point, I was surely capturing a lot of overlap with both the previous and successive scenes – somehow we need to be able to stitch all of these scenes together.

    Try out the Microsoft Live Windows Photogallery, it has a fantastic stitching engine and it is free. The images that it creates also have a special EXIF section which stores parameters necessary to make successful HD View (gigapixel) images – which have support for spherical geometries.

    Would be great if Microsoft could tweak the stitching engine to handle fisheye or equirectangular inputs.

    Seems that the math would be rather complicated though.

  2. eashton
    January 1, 2008 at 10:26 pm

    When I created the composite of input 1 and 2 I did realize that they were not meant to be joined. However, without them being joined, I feel the experience is disjointed, much like a slide show. I suppose, I need to think a little more on how to transition from picture to picture.

    Check out Immersive Media’s demo of their camera system, http://demos.immersivemedia.com/index_directorfound.php?scene=N5
    This doesn’t fit the bill of being free but it is a full hot dog. The question I have is how can we approach something like this but it with cheap/free hardware/software?

  3. eashton
    January 2, 2008 at 1:18 pm

    After a more closer reading of geosquan’s blog and reading the links I think this hotdog is a real probability. From Microsoft’s Photosynth project to Casa’s tools I think we can do this. If I have some time today I’ll explore and post any additional experimenting I do.

  1. No trackbacks yet.

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 )

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s

%d bloggers like this: