The paradigm of active vision advocates studying visual problemsin the form of modules that are directly related to a visual task forobservers that are active. It is argued that in many cases when anobject is moving in an unrestricted manner (translation and rotation) inthe 3D world only the motion's translational components are of interest.For a monocular observer, using only the normal flow-the spatiotemporalderivatives of the image intensity function-the authors solve theproblem of computing the direction of translation. Their strategy usesfixation and tracking. Fixation simplifies much of the computation byplacing the object at the center of the visual field, and the mainadvantage of tracking is the accumulation of information over time. Theauthors show how tracking is accomplished using normal flow measurementsand use it for two different tasks in the solution process. First, itserves as a tool to compensate for the lack of existence of an opticalflow field and thus to estimate the translation parallel to the imageplane; and second, it gathers information about the motion componentperpendicular to the image plane
展开▼