Real Time Recognition of Colored Objects That Is Robust for Illumination Change

Abstract

When color of illumination changes, color recognition by a robot becomes difficult. This paper proposes a method of color recognition that is robust for illumination change. The method consists of three steps. First step is making a look-up table that is robust for brightness change. Second step is estimating color of illumination that can be done real time. Last step is updating the look-up table to adapt to illumination changes. Experiments using the AIBO's camera verify the proposed method.