A 3D sensing technique using fuzzy modeling based on stereo vision

Recently, the 3D sensing technique using plural cameras has been applied to various areas such as visualization, motion capturing, and so on. However, improvement of the camera model calibration is indispensible for more precise measurement. In this study, we propose a fuzzy modeling approach for 3D sensing utilizing the configuration of the stereo vision. A distance between a sensing target and the camera and a distance of an image pixel from the optical axis are used for structuring the fuzzy model for cameras. In our approach, the weighted least mean square error method is successfully applied considering fuzzy partition to formulate the fuzzy model. Through sensing experiments based on the proposed approach, we show effectiveness of the approach in terms of sensing precision.