A method of interactively using a computer to select a sensor combination for analyzing images of a scene in order to distinguish between objects or patterns. Two or more images are acquired, each image from a different sensor. One or more regions of interest (ROIs) are then defined in the scene. A parallel axis plot is then used to display pixel traces of pixels from the ROI, with each axis of the plot representing one of the images. One or more axes may then be constrained, by selecting a range of values on an axis. Pixel traces that satisfy the constraint are illustrated graphically. Further analysis is performed by displaying pixel images of the constrained image, and by performing statistical analysis of the constraint(s).
Michael J. Magee; Ernest A. Franke; Jeremy K. Zoss