A dynamic Jacobian estimation method for uncalibrated visual servoing

Author

Piepmeier, Jenelle Armstrong ; McMurray, Gary V. ; Lipkin, Harvey

Author_Institution

Sch. of Mech. Eng., Georgia Inst. of Technol., Atlanta, GA, USA

fYear

1999

fDate

1999

Firstpage

944

Lastpage

949

Abstract

A dynamic Broyden´s method is presented for use in a quasi-Newton control scheme for model-independent vision guided robotic control. Model independent visual servo control is defined as using visual feedback to control a robot without precisely calibrated kinematic and camera models. The control problem is formulated as a nonlinear least squares optimization. For the moving target case, this results in a time-varying objective function which is minimized using a dynamic Newton´s method. The dynamic Jacobian estimation scheme is used to estimate the combined robot and image Jacobians. Experimental results for a two-degree-of-freedom system demonstrates the success of the algorithm

Keywords

Jacobian matrices; Newton method; feedback; least squares approximations; optimisation; robot vision; servomechanisms; camera models; dynamic Broyden method; dynamic Jacobian estimation method; dynamic Newton method; experimental results; kinematic models; model independent visual servo control; nonlinear least squares optimization; quasiNewton control scheme; time-varying objective function; two-degree-of-freedom system; uncalibrated visual servoing; vision guided robotic control; visual feedback; Cameras; Feedback; Jacobian matrices; Kinematics; Least squares methods; Newton method; Nonlinear dynamical systems; Robot control; Robot vision systems; Servosystems;

fLanguage

English

Publisher

ieee

Conference_Titel

Advanced Intelligent Mechatronics, 1999. Proceedings. 1999 IEEE/ASME International Conference on

Conference_Location

Atlanta, GA

Print_ISBN

0-7803-5038-3

Type

conf

DOI

10.1109/AIM.1999.803299

Filename

803299