Hongming Wang

An adaptive mixture Gaussian background model with online background reconstruction and adjustable foreground mergence time for motion segmentation

Motion segmentation is a very critical task in video surveillance system. In the paper two novel components, background reconstruction and foreground mergence time control, have been incorporated into the adaptive mixture Gaussian background model. The background reconstruction algorithm constructs a static background image from a video sequence that contains moving objects in the scene; then the static background image is used to initialize the background model. The foreground mergence time control mechanism is introduced to make the foreground mergence time adjustable and independent of the models learning rate. Rationales are discussed in detail and experimental results are shown.

Online mapping with a mobile robot in dynamic and unknown environments

In this paper, we address the problem of mapping dynamic and unknown environments. The static and moving objects are modelled as the components in a Gaussian mixture model (GMM). By recursive learning of GMM, the components corresponding to the static objects will have larger weights while the components corresponding to the moving objects will have smaller weights. At each time step, a number of components with the largest weights are adaptively selected as the background map and the new observations which do not match with the background map are classified as the foreground map. In addition, based on a Bayesian factorisation of simultaneous localisation and mapping (SLAM) problem, we present an online algorithm for SLAM with GMM learning. Our contributions are employing GMM learning approach to model the dynamic environment with detection of moving objects and jointing the GMM learning with SLAM in unknown environment. Consequently, an online approach for mapping with a mobile robot in dynamic and unknown environments is presented. Some simulation results indicate that our approach is feasible.

An inversion-free fuzzy predictive control for piezoelectric actuators

Piezoelectric actuators (PEAs) are treated as the core component in the nano-posistioning applications. The inherent hysteresis nonlinearity can dramatically degrade the tracking performance of PEAs. This paper presents an inversion-free fuzzy predictive controller of the parallel distributed structure. A Takagi-Sugeno (T-S) based fuzzy model of PEAs is developed first. With the aid of T-S based fuzzy model, explicit predictive control law can be obtained for each fuzzy rule. Then these predictive control laws are combined by fuzzy inference to generate the overall predictive controller. By the proposed method, the inverse hysteresis model of PEAs is no longer required. A notable feature of the proposed method is that the predictive control law can be obtained before the real-time control of PEAs. Therefore, the on-line computational burden is quite low, leading to a good tracking performance in the high frequency working conditions. Experiments are conducted on a commercial PEA to verify the proposed method. Experiment results show that the proposed method has a satisfactory tracking performance in both the low and high frequency conditions. Comparison results illustrate that the proposed method outperforms some existing approaches such as the inversion-based method and sliding mode control method.

Adaptive Neural Network Tracking Control for Manipulators with Uncertainties

Abstract An adaptive neural network controller is proposed to deal with the end-effector tracking problem of manipulators with uncertainties. By employing the adaptive Jacobian scheme, neural networks, and backstepping technique, the torque controller can be obtained which is demonstrated to be stable by the Lyapunov approach. The updating laws for designed controller parameters are derived by the projection method, and the tracking error can be reduced as small as possible. The favorable features of the proposed controller lie in that: (1) the uncertainty in manipulator kinematics is taken into account; (2) the “linearity-in-parameters” assumption for the uncertain terms in dynamics of manipulators is no longer necessary; (3) effects of external disturbances are considered in the controller design. Finally, the satisfactory performance of the proposed approach is illustrated by simulation results on a PUMA 560 robot.

Scene Analysis for Mobile Robot Based on Multi-Sonar-Ranger Data

The ability of cognition and recognition for complex environment is very important for a real autonomous robot. A new scene analysis method using kernel principal component analysis (PCA) for mobile robot based on multi-sonar-ranger data is put forward. The principle of classification by principal component analysis (PCA), Kernel-PCA, and the BP neural network approach to extract the largest k eigenvectors are introduced briefly. Next PCA, Kernel-PCA and the BP neural network methods are applied in the corridor scene analysis and classification for the mobile robots based on sonar data. At last the experimental results using PCA, Kernel-PCA and the BP neural network are compared and such conclusions are drawn: in common corridor scene classification, the Kernel-PCA method has advantage over the ordinary PCA, and the BP neural network approach can also get satisfactory result.

Mapping Dynamic Environment Using Gaussian Mixture Model

In this paper, we address the problem of mapping dynamic environments with detection of moving objects. The static and moving objects are modeled as the components in a Gaussian mixture model (GMM). By recursively learning of GMM, the components corresponding to the static objects will have higher weights while the components corresponding to the moving objects will have lower weights. At each time step, a number of components with highest weights are adaptively selected as the background map and the new observations which do not match with the background map are classified as the foreground map. In order to obtain the expected observation, from which the Gaussian mixture model is learned, we use a particle filter to approximate the posterior probability density function of the pose of the robot and update it sequentially. Also an on-line algorithm is proposed and some simulations on a simple one-dimensional example indicate that our approach is feasible.

Electron diffraction and lattice imaging study of La2Cu0.925V0.075O4 + d

The superconducting compound La2Cu0.925V0.075O4 +

Sonar Feature Map Building for a Mobile Robot

This paper presents an approach for sonar feature map building. The approach is composed of extracting features at the data-level fusion stage and fusing the extracted features with the registered features in the map at the feature-level fusion stage. A data-level fusion model, termed three measurements association model (TMAM), has been developed for associating three measurements with a line or a point feature. By use of TMAM, different sets of measurements obtained from a single sonar sensor at consecutive steps are associated with the line and point features. Subsequently, the parameters of the identified features are estimated by use of the iterated least square estimation method. Finally, when a feature is extracted, a simple feature-level fusion strategy is used to update the map. The proposed approach has been tested both in simulation and on real data.