A System for Automatic Quantification of Cigarette Smoking Behavior

Cigarette smoking remains the principal source of preventable disease in the United States. Although most tobacco companies have developed filtered cigarettes to reduce the levels of some toxins, smokers have adapted their inhalation patterns to compensate for these newer "light" cigarettes. We describe a new system designed to unobtrusively monitor such elements of smoking behavior. The system used a video camera to capture movies of smoking behavior. Subsequently, image processing and analysis algorithms were applied to the movie frames to form a time-series, termed the smoking topography, which quantified various parameters of smoking behavior