Using range imaging sensors with color imaging sensors in cognitive robot companions: A new and simple calibration technique based on particle swarm optimization

A cognitive robot companion needs sensors in order to enhance its perseptual capabilities. Using different types of sensors in different mechanical settings there is a need to find suitable calibration can be performed between system components or relative to some global reference frame. In this paper we describe a method that can lead to useful calibration models in a generic manner. A particle swarm optimization algorithm is used to optimize a parameter set and suggest methods to reduce or to refine the calibration model iteratively. Relevant equitations do not have to be reformulated as a system of equations and there is no principal restriction on the class of equations. The proposed method is applied to calibrate a new bi-modal sensor head in two different ways: directly between two sensors and to real-world coordinates.