Xie Wei-xin

Online hand-sketched graphics recognition based on attributed relational graph matching

The recognition of online hand-sketched graphics is studied. First, the attributed relational graph (ARG) representation of online hand-sketched graphics is discussed. And then the pattern similarity measure which is computed by using A* algorithm is defined. Finally, a strategy of model matching and a way of model organizing are proposed in order to improve the recognition speed.

New data association technique based on ACO with directional information considered

Abstract Due to the advantages of ant colony optimization (ACO) in solving complex problems, a new data association algorithm based on ACO in a cluttered environment called DACDA is proposed. In the proposed method, the concept for tour and the length of tour are redefined. Additionally, the directional information is incorporated into the proposed method because it is one of the most important factors that affects the performance of data association. Kalman filter is employed to estimate target states. Computer simulation results show that the proposed method could carry out data association in an acceptable CPU time, and the correct data association rate is higher than that obtained by the data association (DA) algorithm not combined with directional information.

Study on capacity of information hiding for still images

Using information theory, a model for the data-hiding channel in a still image is proposed and an evaluation of the number of bits that can be hidden within an image is given. As an example, the number of bits that can be hidden within a standard 512/spl times/512 gray scale Lena image is evaluated.

A new method of building detection from a single aerial photograph

A novel approach based on multifold building features for the task of detecting buildings in aerial photograph is presented. Firstly, a coarse classification based on the texture features was used. Secondly, line adjacency and spectrum features detection are applied to select possible building regions. Thirdly, the fuzzy logic inference is applied to solve the problem of uncertainty and imprecision in building detection. This approach not only has the capability of detecting buildings effectively but also is easily understandable.

Online Visual Multiple Target Tracking by Intuitionistic Fuzzy Data Association

Abstract In this paper, a novel frame-by-frame data association algorithm based on intuitionistic fuzzy sets is proposed for online visual multiple target tracking. In the proposed algorithm, the association costs between targets and measurements are replaced by the intuitionistic fuzzy membership degrees which are obtained by a modified intuitionistic fuzzy c-means clustering algorithm. In addition, in order to mine useful information from the uncertain measurements, a new intuitionistic index is defined and the intuitionistic fuzzy point operator is applied to extract valuable information from the intuitionistic index. Experiments with challenging public datasets demonstrate that the proposed visual tracking algorithm improves tracking performance compared to other algorithms.

Current statistical model based on maximum entropy fuzzy clustering

In the view of the unfitness to the actual maneuver of targets that a fixed maneuvering frequency used in the current statistical model. Firstly, predicted measurements of special maneuvering frequency are clustered with the aid of maximum entropy fuzzy clustering. Then, the estimated means and covariance of the state are mixed by utilizing the fuzzy membership degree of the predicted measurements. Unscented kalman filter is employed to solving the nonlinearity of the measurement equations. Simulation results show that the proposed method has higher accuracy than some existing methods based on the current statistical model in the estimation.

A new method for track initiation in a distributed passive sensor network

A new algorithm is proposed for track initiation in a distributed passive sensor network. The algorithm is formulated using a newly defined fuzzy synthetic closeness function in which the correlations between the set of measurements and targets are reflected. First of all, the algorithm detects targets through searching the globe extreme points of the fuzzy synthetic closeness function using the steepest descent method, then assigns measurements to various targets using threshold test, and finally estimates the initial states of targets using their correlative measurements by Levenberg-Marquart algorithm. The approach does not need any additional information such as the probability of detection, false alarm rate and the clutter density. Simulation results show its effectiveness.

Fuzzy flow detection for IP cut-through

Several solutions, such as multiprotocol over ATM (MPOA) and IP switch router, are given for IP flow transport in ATM networks. Flow detection is necessary in all solutions above to decide when to cut through IP flow. We propose a fuzzy flow detector to give the IP cut-through decision. It shows how to design the detector and how the detector gives the IP cut-through decision. In the fuzzy flow detector system, three input variables are used, i.e., the arrived flow IP packets from the same source, the average interval time between adjacent packets, and the free capacity on the outgoing links to the destination. The output variable is the decision (cut-through or not). Moreover, the hardware implementation is also given.

Fuzzy logic approach to visual multi-object tracking

Abstract In this paper, a novel fuzzy logic data association algorithm is proposed for online visual multi-object tracking. Firstly, in the proposed algorithm, in order to incorporate expert experience into the data association for the improvement of performance in multi-object tracking, a fuzzy inference system based on knowledge is designed by using a set of fuzzy if-then rules. Given the error and change of error of motion, shape and appearance models in the last prediction, these rules are used to determine the fuzzy membership degrees that can be used to substitute the association probabilities between the objects and the measurements (or detection responses). Secondly, in order to deal with the fragmented trajectories caused by long-term occlusions, a track-to-track association approach based on the fuzzy synthetic function is proposed, which can effectively stitch track fragments (tracklets). Because of this, the proposed algorithm has the advantage that it does require no assumption of statistical models of measurement noise and of object dynamics. The experiment results on several public data sets show the efficiency and the ability to minimize the number of fragment tracks of the proposed algorithm.

The combination of ACDA and FCM for Data Association in multi-target tracking

For improving the performance of ACDA(Ant Colony Data Association) for data association in multi-target tracking, we propose the combined method of ACDA and FCM. Since FCM is a determinate algorithm, in nature based on NN(Nearest Neighbor), it could generate reasonable results in any case, which is a backup when ACDA becomes divergent. Experiments have been done in three cases-ACDA as data association method only, FCM only, and the combined method of ACDA and FCM. The results show that the performance of the combined method is superior to the other two methods.