Multi-spectrum field-use image capture & analysis and the RIBG image filter

Abstract

The availability of low-cost wide-spectrum sensors in digital SLR cameras, LED emitters of various wavelengths, and the small form factor of solid state electronics makes it possible to begin to explore the capture of information illuminated by spectrums outside the visible range. It is proposed that the RIBG image filter, where I is infrared in the 700nm - 1000nm range, replace the traditional Bayer pattern. This thesis introduces I-Blending for the creation of traditional 3 channel RGB images from the 4 channel RIBG image. The infrared information is applicable to field capture and analysis because it: (1) limits visual scene disturbance at time of capture, (2) generates multiple images from a single captured image and (3) and provides additional infrared information to exploit. This thesis uses the presence of infrared information in a positional flash sequence algorithm, a flash/no flash algorithm, and 2-D to 3-D surface topology algorithm.