Research On Multi-target Tracking Data Fusion Algorithms Based On Rough Set

Multi-target tracking has gained popularity over past decade with advent of vigorous sponsorship of many areas. Multi-target tracking technique is depended on data fusion technique. Data fusion is a burgeoning field of information processing technology. Although it has been developed greatly, it has not formed an integrated area, and still has many problems to solve. Many relevant theories and techniques about multi-target tracking also remain to be developed for practice.This dissertation expatiates on the principle, definiens, structural model, algorithms and application of data fusion. This dissertation also applies a survey in the data fusion research field. This dissertation combines the advantages of fuzzy set theory, neural network and rough set, and constructs a multi-target tracking frame based on rough set and fuzzy neural network. The main contents of this dissertation is as following:(1) According to the rationale of fuzzy set theory, this dissertation provides a fuzzy reasoning target tracking algorithm based on Kalman filter.(2) This dissertation designs a multi-target tracking method, which combines Kalman filter with fuzzy neural network to form a closed loop.(3) By combining rough set with fuzzy neural network, this dissertation proposes a multi-target tracking data fusion algorithm based on rough set and fuzzy neural network. With a simple exemplification, this dissertation proves the algorithm is right.