Since Stanley Kubrick's, epic 1968 science fiction film, "2001: A Space Odyssey", computer comprehension of visual human speech directly to words has been imagined. There have been hundreds of publications and book chapters devoted to this topic, including multiple review articles. Examples are Gaikwad et al. [1] and Battista et al. [2]. Most papers depend on Artificial Intelligence based upon Hidden Markov Models and other camera techniques capturing and analyzing visual motion in time of video frames. Some notable exceptions are direct monitoring of the laryngeal, Jorgensen [3] and embedding magnets, Gilbert et al. [4]. In this paper, a novel technique is used to improve the capture of these motions and process the utterance of an entire word into a Composite Ensemble Phonemes Viseme (CEPV). Described is the improvement in the optical sampling technique. This is accomplished both with an active optical source and a high speed camera. Subsequent digital signal processing allows for compression of time domain scattering of light of an utterance into a unique image or matrix of numbers, representing a word which can be compared to other images within a list to interpret meaning.

Lip Reading Discrete Fourier Transform Laser Illumination Covert Communications Urban Security/Warfare