Research On Target Association Method Based On Remote Sensing Information

The tactical application about the remote sensing data is a complicated system engineering. Target association analysis is one of the crucial problems to be solved in the tactical application.The classical target association algorithms are analyzed and their merits and demerits are discussed. Considering the low time resolution of the remote sensing data, the battlefield target association algorithms are studied with the target attribute and all kinds of accessorial information.Three aspects are studied in this thesis:1. A target association model based on target attribute is proposed in this thesis. Firstly the characteristic of remote sensing and accessorial information is introduced. Then a target association model is given. To improve target association accuracy, a target association factor matrix is derived with the reasoning mechanism of the fuzzy expert system.2. A target association algorithm based on the fuzzy similarity measure is discussed. Remote sensing data may be regarded as fuzzy vector because it isn't accurate. When fuzzy vector is founded by fuzzed data, the targets will be distinguished with the similarity measure matrix of the fuzzy vector.3. A target association algorithm based on Bayesian theory is put forward. The target association may be considered as a classing process. With the Bayesian classing model as the judgment basis, the current battlefield target will be distinguished by way of above Bayesian classing model.In the end, it is pointed out that some problems are to be solved in the next work.