From data-flow task to multitasking: applying the synchronous approach to active vision in robotics

Abstract

We apply the synchronous approach to real time active visual three dimensional (3D) reconstruction using a camera mounted on a robot end effector. It illustrates the adequateness of SIGNAL, a synchronous data flow programming language and environment, for the specification of a system dealing with various domains in central theory and computer vision. More precisely, our application consists in the 3D estimation of a set of geometrical primitives using an active vision paradigm. At the level of camera control, the visual servoing approach is specified and implemented in SIGNAL as a function from sensor inputs to control outputs. Furthermore, the 3D estimation method is based on the structure from controlled motion approach (constraining camera for optimal estimation). Its specification is made in parallel to visual servoing, and involves the delay mechanism of SIGNAL for the specification of filters. This reconstruction involves to be focused on each object; we thus present a perception strategy for sequencing several estimations, using task preemption and time intervals in SIGNAL. It allows us to consider in an unified framework the various aspects of the application: from dataflow task specification to multitasking and hierarchical task preemption. The integration of these techniques is validated experimentally by their implementation on a robotic cell. Merits and drawbacks of the proposed framework with respect to more typical asynchronous approaches are also discussed.