Data fusion system based on dissimilar sensors;

To improve calculation performance of centralized target tracking system which consists of multiple dissimilar sensors, a fusion tracking method based on equivalent sensor was presented.

In order to assure the parameter reliability and veracity of collecting datum, according to the characteristics and structure of the system, three data fusion methods are proposed: same sensor data fusion with different measurement periods, multi-sensors data fusion for identical sensor, dissimilar sensors data fusion.

Hough transform track initiation algorithm for distributed heterogeneous sensor network;

Fast target location method in the heterogeneous sensor system;

Based on linear unbiased minimum variance estimation theory, a fusion algorithm which fused the state vector of nonlinear systems with dissimilar sensors with arbitrary correlated noises is developed.

Illuminate the characteristic of distributed dissimilar sensor network,aimed at the multiformity and fuzzy of the information in the network.

A new kind of traffic sensor which made from is introduced in this paper,from the work theory we can find that this new traffic sensor can be used for collecting many signals such as number of axles,distance between axles,vehicle speed,width between wheels,number of tires and load on axles,it is widely used in vehicle classification,weigh in motion,tollgate,red light camera…