Computer programs and computer-implemented methods are disclosed for
extracting and analyzing combinatorial chemical data from images. The
invention receives a stream of data, identifies a plurality of regions of
interest and determines a series of values for one or more regions of
interest. The library includes a plurality of members. The data stream
includes a series of images of the library. Each region of interest
corresponds to a location in each of the images and to a location in the
library. The series of values for each region includes a value for each of
the images. Particular embodiments have the following features. The
invention calculates a figure of merit for a library member from each
series of values. The stream of data comprises a series of images
generated at a frequency of greater than about 1, 6, 12 or 30 frames per
second. The series of images is captured from a beginning of a
combinatorial experiment to an end of a commercial experiment. The regions
of interest are identified after the series of images has been captured.
The images are infrared images. The figure of merit is an emittance
change, a phase transition point or a thermoelectric figure of merit for a
library member. Regions of interest are identified by a user and have a
user-defined shape. The series of values for each region is determined
based on a user-defined computation. A graphical representation of the
figures of merit is displayed.
