Using Optic Flow for the Estimation of Travel Distance

Much research on optic flow has been concerned with the estimation of heading and the control of the direction of self-motion. In my presentation I will instead focus on a different use of optic flow, namely the estimation of the distance that one has traveled. Optic flow in itself does not provide travel distance, only the combination of distance and speed (time-to-contact) is directly available. However, when scaling information from the environment is present, such as the distance to the ground when standing or walking on flat terrain, an integration of the optical velocity can yield ego-speed and travel distance. I will present experiments that show that humans can use optic flow for the estimation of travel distance, but that they often under- and, in some conditions, overestimate a movement's extent. I will then present a model of travel distance estimation from optic flow that is based on leaky path integration.