Search

Surveillance on pedestrians or other moving objects in crowded sites is a crucial and challenging problem for various security tasks. To solve this problem, this paper addresses two energy methods which are based on intensity variation and motion features respectively. With wavelet analysis of the energy curves, the result shows that both methods can be used to deal with crowd modeling and real-time surveillance satisfactorily. A comparison of the two methods is then made in the real environment in a metro surveillance system.

For more precise control of the FES-derived gait, the timing of specific gait phases is needed. With the information of when the legs are in each phase of gait, the quality of the gait during each phase can be assessed. The goal of this paper is to investigate potential use of machine learning techniques, in particular Support Vector Machine (SVM), and Force Sensitive Resistor (FSR) to detect discrete stages in the cyclic motion of dynamic human gait. We present intelligent shoes for capturing and analyzing dynamic human gait data. Principal Component Analysis(PCA) will be applied for feature generation and data reduction, and Support Vector Machine (SVM) will be applied for training and classifier generation. The classification performance is assessed quantitatively, by measuring the percentage of time steps in which the correct event is found and qualitatively, by observing the types of errors encountered. The experimental results verify that the proposed method is valid and useful with overall accuracy to 91%.

In this paper, the topological structure optimization problem in a wireless sensor network is formulated as a graph problem, while fulfilling random node-to-node communication demands. We propose a new optimization method: un- detour optimization, which is a kind of improved genetic algorithm. Base on this method, a topological structure optimization plan for a given sensor network can be built. We evaluate this approach quantitatively by simulating community wireless sensor networks and comparing it with other approaches. We demonstrate that significant performance advantages can be achieved by this approach.

The capturing operation for the space object by space robot can not be avoided in order to complete the on- orbital services, such as repairing, maintaining, inspecting and exchanging orbital replace unit (ORU). However, there are some unknown factors, such as the unknown payload and unknown modeling of object and the other uncertain dynamic parameters in the space robot, will affect the performance and the security of the space robot when the space robot captures the unknown object. Therefore, it is significant to design a robust control law that will guarantee the performance and security of the space robot under these uncertainties. Here, we propose a new optimal control method to robust control of space robot. To design the robust control law, we translate the robust control problem into an optimal control problem, where the uncertainties are reflected in the objective function index. We optimize this objective function to realize the robust control. An illustrative example shows that the proposed approach is remarkable.

Automatic recognition of human actions is an important but difficult problem in the area of computer vision. In this paper, an novel approach is introduced to handle the problem. Firstly, human body is detected through the use of contour information and the body is tracked by the hybrid method. The most prominent features are searched by using the mean-shift method based on the body structure and the history motion image information. Finally, a learning method based on the multiple support vector machines is used to learn action types dynamically. We propose a method which integrate the Bayesian framework with the SVM method, which largely improves the recognition rate using historic information. Experiments show that our system can run in realtime for the detection of abnormal behaviors with limited information and produces robust result by making full use of historic motion information.

This paper presents a method for modeling human abnormal gait using hidden Markov model under the framework of a shoe-integrated system. The intelligent system focuses on modeling the following patterns: normal gait, toe in and toe out gait abnormalities. In the developed prototype, an inertial measurement unit (IMU) consisting of three-dimensional gyroscopes and accelerometers is employed to measure angular velocities and accelerations of human foot. Four force sensing resistors (FSRs) and one bend sensor are arranged on a insole of each foot for force and flexion information acquisition. The proposed method is mainly based on principal component analysis (PCA) for feature generation and hidden Markov model (HMM) for multi-pattern modeling. The "similarity distance measure" criterion is introduced to do model-to- model evaluation. Experiment results demonstrate the proposed approach is robust and efficient in detecting abnormal gait patterns. Our goal is to provide a cost-effective system for detecting gait abnormalities in order to assist persons with abnormal gaits in developing the normal walking pattern in their daily life.