Comparing information content in SWIR, eSWIR, MWIR, superband, and multiband-fused images

The two-dimensional (2-D) correlation coefficient calculation is a method for comparing the similarity between two data sets or images. This method was used to quantify the complementary information between images taken in the short-wave infrared (SWIR), extended SWIR (eSWIR), mid-wave infrared (MWIR), SWIR through MWIR superband, and two multiband-fused images—one with SWIR, eSWIR, and MWIR bands, and the other with only SWIR and MWIR bands. Data collections were performed at various locations of both rural and urban scenes using a 1–6 µm superband imaging system with (1) a 1.9 µm SWIR short-pass filter, (2) a 2–2.5 µm extended SWIR (eSWIR) bandpass filter, (3) a 3 µm MWIR longpass filter, and (4) no filter for the full superband response. For each individual scene, the correlation coefficient between each pair of the six images was determined, representing the amount of shared information in that scene between the two bands being compared. This process was repeated 11 times for 10 daytime image sets and one nighttime image set of both rural and urban scenes. The results for all 11 data sets were compiled to determine the average correlation coefficients in all six bands for both types of scenes.