针对姿态多变化的飞机自动目标识别中的低识别率问题,提出了一种基于DSm T(Dezert-Smarandache theory)与隐马尔可夫模型(Hidden Markov model,HMM)的飞机多特征序列信息融合识别算法(Multiple features and sequential information fusion,MFSIF).其创新性在于将单幅图像的多特征信息融合识别和序列图像信息融合识别进行有机结合.首先,对图像进行二值化预处理,并提取目标的Hu矩和轮廓局部奇异值特征;然后,利用概率神经网络(Probabilistic neural networks,PNN)构造基本信度赋值(Basic belief assignment,BBA);接着,利用DSm T对该图像的不同特征进行融合,从而获得HMM的观察值序列;再接着,利用隐马尔可夫模型对飞机序列信息融合,计算观察值序列与各隐马尔可夫模型之间的相似度,从而实现姿态多变化的飞机目标自动识别;最后,通过仿真实验,验证了该算法在飞机姿态发生较大变化时,依然可以获得较高的正确识别率,同时在实时性方面也可以满足飞机目标识别的要求.另外,在飞机序列发生连续遮挡帧数τ≤6的情况下,也具有较高的飞机目标正确识别率.
Most of modern systems for information retrieval, fusion and management have to deal with more and more qualitative information （by linguistic labels） besides information expressed quantitatively （by numbers）, since human reports are better and easier expressed in natural language than with numbers. In this paper, Herrera-Martfnez＇s 2-Tuple linguistic representation model is extended for reasoning with uncertain and qualitative information in Dezert-Smarandache Theory （DSmT） framework, in order to overcome the limitations of current approaches, i.e., the lack of precision in the final results of linguistic information fusion according to 1-Tuple representation （ q1 ）- The linguistic information which expresses the expert＇s qualitative beliefs is expressed by means of mixed 2 Tuples （equidistant linguistic labels with a numeric biased value）. Together with the 2-Tuple representation model, some basic operators are presented to carry out the fusion operation among qualitative information sources. At last, through simple example how 2-Tuple qualitative DSmT-based （q2 DSmT） fusion rules can be used for qualitative reasoning and fusion under uncertainty, which advantage is also showed by comparing with other methods.
With the increment of focal elements number in discernment framework,the computation amount in Dezert-Smarandache Theory (DSmT) will exponentially go up. This has been the bottleneck problem to block the wide application and development of DSmT. Aiming at this difficulty,in this paper,a kind of fast approximate reasoning method in hierarchical DSmT is proposed. Presently,this method is only fit for the case that there are only singletons with assignment in hyper-power set. These singletons in hyper-power set are forced to group through bintree or tri-tree technologies. At the same time,the assignments of singletons in those different groups corresponding to each source are added up respectively,in order to realize the mapping from the refined hyper-power set to the coarsened one. And then,two sources with the coarsened hyper-power set are combined together according to classical DSm Combination rule (DSmC) and Proportional Conflict Redistribution rule No. 5 (PCR5). The fused results in coarsened framework will be saved as the connecting weights between father and children nodes. And then,all assignments of singletons in different groups will be normalized respectively. Tree depth is set,in order to decide the iterative times in hierarchical system. Finally,by comparing new method with old one from different views,the superiority of new one over old one is testified well.
