Active stereo vision based system for estimation of mobile robot orientation using affine moment invariants

The computation of a mobile robot position and orientation is a common task in the area of computer vision and image processing. For a successful application, it is important that the position and orientation of a mobile robot must be determined properly. In this paper, a simple procedure for determining the orientation of a mobile robot using two cameras is presented. The two cameras are used to capture the images of a mobile robot at various orientations. Four simple neural network models are developed to associate the inputs and output (orientation). First and second neural network modes are used to estimate the orientation of a mobile robot using only the features derived from the first and second camera respectively. The third neural network model is used for estimating the orientation of a mobile robot using features derived from both the cameras. Finally, the fourth neural network model is used to estimate the orientation of a mobile robot using features derived from the combined image. Simulation results show that the proposed algorithm can be used to estimate the orientation of the mobile robot accurately.

[1]  Hüseyin Ekiz,et al.  A study of neural network based inverse kinematics solution for a three-joint robot , 2004, Robotics Auton. Syst..

[2]  Helder Araújo,et al.  Maintaining the relative positions and orientations of multiple robots using vision , 2001, Pattern Recognit. Lett..

[3]  M Paulraj.  Introduction to Artificial Neural Networks , 2003 .

[4]  Jan Flusser,et al.  Landmark recognition using invariant features , 1999, Pattern Recognit. Lett..

[5]  Tony Szturm,et al.  Application of feedforward backpropagation neural network to center of mass estimation for use in a clinical environment , 2003, Proceedings of the 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (IEEE Cat. No.03CH37439).

[6]  Jovisa D. Zunic,et al.  Boundary based shape orientation , 2008, Pattern Recognit..

[7]  Ming-Kuei Hu,et al.  Visual pattern recognition by moment invariants , 1962, IRE Trans. Inf. Theory.

[8]  Stephen M. Rock,et al.  Experiments in Control of a Free-Flying Space Robot Using Fully-Connected Neural Networks , 1999 .

[9]  Margrit Betke,et al.  Mobile robot localization using landmarks , 1997, IEEE Trans. Robotics Autom..

[10]  J. A. Janet,et al.  Autonomous Mobile Robot Global Self-Localization Using Kohonen and Region-Feature Neural Networks , 1997 .

[11]  Jan Flusser,et al.  Rotation Moment Invariants for Recognition of Symmetric Objects , 2006, IEEE Transactions on Image Processing.

[12]  Stephen M. Rock,et al.  Gradient‐based parameter optimization for systems containing discrete‐valued functions , 2002 .

[13]  Tien-Fu Lu,et al.  An on-line relative position and orientation error calibration methodology for workcell robot operations , 1997 .