Spaced Out :: 04 2010

I was recently introduced to Prezi, a rapid presentation package for those sick of the usual Powerpoint interface. It is programmed in Adobe AIR and so requires Flash. There is a free version which has limited storage and makes presentations freely available. A personal version (which is free to teachers and students) and a professional version (which gives the option of an offline app). It broadly tales the paradigm of a mindmap (and for those interested, FreeMind is an excellent open source implementation) and lets you type text and then add further ideas around it. It doesn’t use branches though, rather just allowing you to dump text, images, audio and video on to the “canvas”. What is unique is the zooming concept which presents unlimited zoom levels and allows you to nest mindmaps within mindmaps. When it comes to finalising the presentation you add “paths” for the presentation to follow and zoom in and out to each area. Formatting is very limited amounting to colour for text, size, rotation and boxes. Thats about it. But the power lies in the simplicity and speed. I’m doing my first talk with Prezi tomorrow so will see what the audience thinks; its certainly a paradigm shift to the speaker as it bypasses the whole “dump everything on to a slide” so you have to spend more time preparing.

Paul Beaty has an interesting comparison of the effect of multiple processors when using ERDAS Imagine for processor intensive tasks. In his testing he did two large exports using ECW and JPEG2000. Now I’ve always found Imagine to be pretty robust and fast; the algorithms appear well written, stable and designed to complete efficiently (although I can’t comment on the 2010 version as the disc is currently sitting on my desk unused). Perhaps harking-back to the days when raster processing was costly in computing terms (and so monetarily as well). Anyway, the comparison is stark and clearly 2 and 3 cores are where the biggest gains are made, but if you do this alot then the more cores the merrier!

I blogged last year on the increasing use of Python as the preferred language for geospatial automation driven, in no small part, but ESRIs uptake. Anyway, a useful post on essential Python modules for the geospatial programmer.

Plenty of Eyjafjallajökull stuff in the blogs at the moment (not surprisingly!) so I thought I would compile a few remote sensing bits together:

MODIS RapidFire had one of the earlier sets of imagery of the ash cloud as it spread out over Europe. The high spatial resolution and twice a day imaging makes it very good for this type of stuff.

Robert Peston provides a nice summary. Couple of nice quotes: “[This] shows that the issue isn’t whether the cloud is real and dangerous - but whether its extent can be accurately mapped.” Spatial extent is clearly the most important, but vertical mixing is rapidly becoming a key issue. And, of course:

“Right now, the biggest impact for business is the sheer number of executives who are stuck abroad, unable to come home. ‘The real danger for them is that we’ll discover we don’t really need them,’ one business leader joked.” Nice touch!

The Map Room have taken an animated GIF of the spread of the ash cloud as produce by the Norwegian Met Office and put it on YouTube.

Lidar News has covered how Doppler LiDAR is being used to determine vertical extent.

A nice series of images over at Earth Observatory

So, plenty going on in the remote sensing world.

The Geoprocessing blog over at ESRI highlights an interesting (well, in an academic sense!) book on spatial statistics which they have contributed a chapter to on ArcGIS. The chapter has been made available for download so is well worth checking out.