Seeing through the Fog: An Algorithm for Fast and Accurate Touch Detection in Optical Tabletop Surfaces

Abstract

Fast and accurate touch detection is critical to the usability of multi-touch tabletops. In optical tabletops, such as those using the popular FTIR and DI technologies, this requires efficient and effective noise reduction to enhance touches in the camera's input. Common approaches to noise reduction do not scale to larger tables, leaving designers with a choice between accuracy problems and expensive hardware. In this paper, we present a novel noise reduction algorithm that provides better touch recognition than current alternatives, particularly in noisy environments, without imposing higher computational cost. We empirically compare our algorithm to other noise reduction approaches using data collected from tabletops at research labs in Canada and Europe.