PERSPECTIVE SYSTEM을 위한 비선형 저차화 관측기 설계와 영상 기반 VISUAL SERVOING 적용에 대한 연구

Abstract

In this thesis we propose a nonlinear reduced order observer design method for estimating the depth of an object feature using images obtained by a single camera. A nonlinear output injection function is introduced to design a first order reduced observer that has only one gain parameter. The estimation error of this observer converges asymtotically to zero. We also present a simple selectionprocedure for the gain parameter that satisfies the convergence condition. Simulation results show the efficiency of the proposed observer and the validityof the proposed observer was verified by an experimental application on camera in robot hand. The proposed nonlinear reduced observer is applied to image based visual servoing in order to estimate the image jacobian. In the image-based visual servoing framework, error signals are directly computed from image feature parameters, thus obtaining control schemes which do not need neither a 3-D model of the scene, nor a perfect knowledge of the camera calibration matrix. However, the current value of the depth for each considered feature must be known. We propose a method to estimate on-line the value of Z for point features while the camera is moving through the scene, by using nonlinear reduced order observer. By interpreting Z as a continuous unknown state with known dynamics, we build an estimator which asymptotically recovers the actual depth value for the selected feature. Image based visual servoing using nonlinear reduced order observer is simulated, and the simulation results show the necessity of the proposed observer, and the efficiency of the proposed observer.