One of my favorite features that has been added to ENVI in the past several years is the Feature Extraction workflow. That tool segments an input image into polygonal objects, then classifies those objects into categories using the spatial, spectral, and/or texture attributes that it calculates for each object.

So, what I’m most excited for in the new ENVI 5 release are some improvements that will make the Feature Extraction workflow even better. For one thing, new high-speed multi-threaded processing makes the performance much faster. Anyone else who has waited a couple of days for the processing on a large image to finish will understand my enthusiasm for this change. I’m told that in some cases, the segmentation can be many times faster than it is in ENVI 4.8.  That’s pretty impressive, and very welcome.

Also, I really like the changes to the interface, which place the segmentation and merge parameters, as well as some other preprocessing parameters, on the same initial dialog. This change means that you can play around with interactions of the various parameters, checking out the results in the instant preview. Then when you’ve figured out the optimal settings, you can run the segmentation and merge steps all in one shot. That is going to save me a bundle of time right there.

Finally, it’s also now possible to output some intermediate steps, such as the attribute images. Those images contain a whole lot of information, and I can imagine that people will come up with all kinds of alternate ways of using them outside of the rest of ENVI’s Feature Extraction workflow. As long as I have your attention, I also want to mention that the Exelis VIS Tech Support team has been working hard writing Help Articles to help people find answers to ENVI 5 questions that we think may be common. So, if you run into any specific technical questions about ENVI 5, you might try checking the Help Article database, to see if your question has already been answered.

The new interface for the feature extraction workflow in ENVI 5.