Genie Pro®

The original satellite image of the Los Alamos National Laboratory next to the final product of a 12 land-cover class Genie Pro® image of the same area.

Genie Pro® 2.0 is an adaptive automatic feature extraction (AFE) software application that can be used to search and classify remotely sensed imagery.  The software uses techniques from statistical machine learning theory and evolutionary computation theory – also known as genetic programming – to perform robust and customized AFE in multispectral, hyperspectral, panchromatic, and multi-instrument fused imagery.

Genie Pro® features an easy-to-learn-and-use graphical user interface, with intuitive tools for selecting training regions and editing results.  Genie Pro® uses the training data to evolve algorithms based on a set of spectral and spatial image processing operators – it essentially writes an image processing algorithm for you.  Genie Pro® represents a significant technological advance over standard image classification algorithms currently used in GIS and image processing packages.  It can be used to distinguish between similar features that share spectral characteristics – for example, asphalt roads and asphalt roofs.  The program considers the combined spatial and spectral context of each pixel, enabling it to create accurate map overlays for tasks that would be the downfall for traditional spectral-only classification algorithms.

The most convenient aspect to Genie Pro® is that is requires no previous programming experience to use. It captures the skills of an image scientist into a simple step-by-step process and is easy to comprehend.

System Requirements

Genie Pro® 2.0 works on both Microsoft Windows® operating systems (2000, XP, Vista) and Macintosh Mac OS X (Requires Mac OS X v. 10.4 or later).