Simultaneous estimation of optical flow and heat transport in infrared image sequences

Abstract

This paper incorporates physical models of heat transport into motion analysis in infrared image sequences. Physical transport processes, such as heat diffusion and decay, are causing time dependent brightness variations, violating the common brightness constancy assumption. Previous approaches to optical flow computation have accommodated violations of brightness constancy with the use of robust statistics or with generalized brightness constancy constraints that allow generic types of contrast and illumination changes. Here, we consider realistic models of brightness variation that have time-dependent physical causes. We simultaneously estimate the optical flow and the relevant physical parameters, such as the heat diffusion and decay constants. The estimation problem is formulated for a wide class of physical models using total least squares (TLS), with confidence bounds on the parameters.