Hao Wan

Novel two-stage hybrid IBVS controller combining Cartesian and polar based methods

In this paper, we address the issue of two-stage visual servoing using pure 2D visual data. For robotics visual control, Imaged based visual servoing(IBVS) is considered more efficient and robust. Classical IBVS system is usually defined in the Cartesian coordinate space as image points. It is known that classical IBVS may cause unexpected camera movements for cases involve large rotation motion around optical axis, while IBVS based on polar coordinate can overcome this problem. We made a further study on polar-based IBVS control, and proposed a two-stage strategy combining both basic IBVS methods. Based on task decomposition using information from homography, we designed a novel two-stage controller considering strengths of both basic IBVS systems. Simulations are taken to compare with the basic IBVS methods. Finally, the results are presented to illustrate the effectiveness and optimal performance of the proposed method.

Novel hybrid IBVS based on the unified projection model

In this paper, a novel hybrid IBVS method is proposed to improve the performance of robotics visual control system. In this work, we only consider image features to define the error function. Based on the unified projection model, features are selected to control full-DOF motion of the robot. Spherical projected features are used to control the translational motion, while perspective projected features are used to control the rotational motion. These hybrid projected features ensure nice decoupled properties in control system. In all, two sets of features are proposed to design the IBVS controller. One is based on homography-based features, the other is based on moment invariants features. Simulation results are provided to illustrate the effectiveness and optimal performance of the proposed method.

Image based visual servoing from hybrid projected features

in this paper, we address the issue of hybrid visual servoing, which improve the performance of robotics system. Different from most of existing hybrid methods combining image based(IBVS) and position based(PBVS) visual servoing methods, we consider directly image features in error function, which is a robust IBVS method. In the framework of unified projection model, we select six proper features to control full-DOF motion. Invariants to rotational motion are computed with spherical projection model and used to control the translational motion, while invariants to translation motion are computed with perspective projection model and used to control the rotational motion. These hybrid projected features ensure good decoupled control between translational and rotational motions. Furtherly, two virtual moments calculated with SVM regression model are proposed to control the rotational motions around the x-axis and y-axis, which ensure improved decoupled properties. Simulation results are provided to illustrate the effectiveness and optimal performance of the proposed method.

Feature Selection Tracking Algorithm Based on Sparse Representation

In order to enhance the robustness of visual tracking algorithm in complex environment, a novel visual tracking algorithm based on multifeature selection and sparse representation is proposed. In the framework of particles filter, particles with low target similarity are first filtered out by a fast algorithm; then, based on the principle of sparsely reconstructing the sample label, the features with high differentiation against the background are involved in the computation so as to reduce the disturbance of occlusions and noises. Finally, candidate targets are linearly reconstructed via sparse representation and the sparse equation is solved by using APG method to obtain the state of the target. Four comparative experiments demonstrate that the proposed algorithm in this paper has effectively improved the robustness of the target tracking algorithm.

New results in visual invariant features for improved image based visual servoing

Features selection plays an important role in image based visual servoing, the objective is to determine six independent visual features with nice decoupled and linear properties. Visual invariant is well-known as an ideal kind of features to design an improved IBVS. In this paper, a kind of visual invariant to rotational motions is derived from projection onto surface of a unit sphere, and it is calculated by vector inner product operation. Invariant based feature jacobian can be provided through a simple deduction with operation laws of vector or matrix. Then, six independent visual features are determined with three inner product features and homography-based features, and a decoupled IBVS controller is proposed. Finally, simulations results with a free-flying perspective camera are provided to illustrate the significance of our proposed method.

Uncalibrated visual predictive control using Broyden estimation

in order to increase the flexibility and intelligence of robotics visual control system, Broyden estimation of composite Jacobian is combined with nonlinear model predictive control in robotics visual servoing. This method does not require the knowledge of robot model, camera model and object model, while visibility constraints, limitations of joint angles and joint velocities are all considered. Robotics control is achieved by solving the constrained optimization problem over a predictive horizon at each step, corresponding predictive model is designed with estimated composite Jacobian. Finally, simulation results are provided to demonstrate the validity of proposed method.

Visual predictive control from distance-based and Homography-based Features

This paper deals with the issue of improved performance for image based visual servoing combining features selection and constraints handling. Observed object is described as a set of 3D points, corresponding 2D pixel points features in camera view are extracted and transformed to spherical point features through a virtual projection process; then, a set of six independent visual features combining distance-based and Homography-based Features, is determined and the related feature Jacobian is calculated. After features selection, image based visual servoing (IBVS) is formulated into a constrained optimization problem by nonlinear model predictive control (NMPC) method. Visibility constraints, 3D velocities constraints and 3D task space constraints are all considered in the optimization problem. Finally, simulations are provided and demonstrate the effectiveness and improved behavior of proposed method.

Nonlinear model predictive control for visual servoing from a set of points

From a set of points in the image plane, six independent image features are designed with nice decoupling and linear properties. Then, nonlinear model predictive control is used to redefine the visual servoing as a constrained optimization problem, where local predictive model is designed based on image jacobian. The necessarity of field-of-view constraint is deceased, thanks to proposed strategy for recovery of disappearing matched point features by projection homography. This helps to reduce the unknown parameters in the optimization problem. On the other hand, proposed constrained visual servoing control is expected to perform well in task space because of selected features. Finally, simulation results are presented to illustrate the effectiveness and improved performance of proposed method.

Implementing improved repetitive control in printing and die-cutting machine servo system for disturbance rejection

In this paper, we address the issue of improved performance of servo control for corrugated paper printing and die-cutting machine. For two important operations in corrugated paper manufacture process (lead-edge feeding and die-cutting), the motion characteristics are analyzed. Improved repetitive control method is proposed to improve accuracy of the system by rejecting disturbance. The improved method combines sliding mode control and repetitive control, repetitive control is used to reduce or eliminate periodic disturbances while sliding mode control is used to reduce the impact of load change. To verify the effectiveness of proposed control method, we take feeding roller servo system for lead-edge feeding operation as an example. According to the structure of lead-edge feeding servo system, an identification experiment is designed for feeding roller servo system, and obtain the mathematical model of the system by analyzing data of frequency domain response, then simulation results are present to illustrate the improved performance of proposed method.