A Robust Feature Extraction Method for Image-Based Visual Servoing

This paper deals with the visual positioning task in the image-based visual servoing. The controller is based on the discretized model of image Jacobian matrix. The visual servo controller can solve the intractable problems, such as large displacements between the initial and the desired pose of the camera. This controller can achieve a smooth and linear image trajectory in the image space. The developed IBVS controller also enhances the camera trajectory in 3-D space with time-varying depth. The proposed line feature extraction method is robust to the image noises, the dim light condition and the shadows. The simulation test is performed to validate the effectiveness of the proposed image-based visual controller and line feature extraction method.