基于动态时间规整的空中目标机动识别
|
摘要
针对现有空中目标机动识别方法识别率较低且耗时较长的问题,提出一种基于动态时间规整(Dynamic Time Warping,DTW)的空中目标机动识别方法。对目标跟踪产生的空中目标飞行轨迹进行机动序列提取,并将其作为测试模板,用动态时间规整算法匹配测试模板和参考模板,以识别空中目标的机动类型。最后基于相同的训练样本和测试库,在相同拒识率下,将DTW算法与隐马尔科夫(Hidden Markov Model,HMM)算法进行识别效果对比。实验结果表明,相较于HMM算法,所提算法在低训练样本条件下的识别率较高,且识别时间更短。
Current methods in recognizing the aerial maneuvering target have a low recognition rate and much time consuming. Aiming at the problem,an aerial maneuvering target recognition algorithm based on dynamic time warping(DTW) is proposed. Firstly,it extractes the dynamic sequence of the aerial target flight trajectory generated by the target tracking,and uses it as a test template. Then,match the test template with the reference one using the DTW algorithm,so as to recognize the maneuvering type of aerial maneuvering target. Finally,based on the same training samples and test base,DTW algorithm with the hidden Markov model(HMM) under the same rejection rate are compared. The comparison results reveals that DTW algorithm has a higher recognition rate and spends less time as well.
Current methods in recognizing the aerial maneuvering target have a low recognition rate and much time consuming. Aiming at the problem,an aerial maneuvering target recognition algorithm based on dynamic time warping(DTW) is proposed. Firstly,it extractes the dynamic sequence of the aerial target flight trajectory generated by the target tracking,and uses it as a test template. Then,match the test template with the reference one using the DTW algorithm,so as to recognize the maneuvering type of aerial maneuvering target. Finally,based on the same training samples and test base,DTW algorithm with the hidden Markov model(HMM) under the same rejection rate are compared. The comparison results reveals that DTW algorithm has a higher recognition rate and spends less time as well.
引文
[1]冷画屏,吴晓峰,王慕鸿.空中目标战术机动类型的实时识别[J].火力与指挥控制,2011,36(1):64-66.
[2]童奇,李建勋,童中翔,等.基于机动识别的空战意图威胁建模与仿真[J].现代防御技术,2014,42(4):174-184.
[3]王利芬,李莉,聂志强.超视距空战中目标机动意图评估[J].电光与控制,2012,19(12):68-71.
[4]谢川,倪世宏,章宗麟,等.一种基于知识的特技飞行动作快速识别方法[J].计算机工程,2004,30(12):116-118.
[5]尹文军,倪世宏.一种基于遗传算法的飞行动作识别方法[J].电脑开发与应用,2006,19(4):50-51.
[6]张军,陈付彬,付强,等.基于HMM的机动目标识别[J].国防科技大学学报,2003,25(2):51-55.
[7]宋元,王永春.海上防空作战态势估计理论及应用[M].北京:国防工业出版社,2014:100-126.
[8]ANDERSSON M.Classification of aerial missions using hidden markov models[C]//Proceedings of the 7th European Conference Symbolic and Quantitative Approaches to Reasoning with Uncertainty,2003:125-136.
[9]RABINER L R. A tutorial on hidden markov models and selected applications in speech recognition[J].IEEE Transactions on Acoustice,Speech and Signal Processing 1989,77(2):257-286.
[10]王芳,吴志泉,史红权. SVM在空中目标威胁值评估中的应用[J].火力与指挥控制,2017,42 (9):30-33.
[11]覃鸿,王守觉.多权值神经元网络仿生模式识别方法在低训练样本数量非特定人语音识别中与H MM及DTW的比较研究[J].电子学报,2005,33(5):957-960.
[12]SAKOE H,CHIBA S. Dynamic programming algorithm optimization for spoken word recognition[J]. IEEE Transactions on Acoustice,Speech and Signal Processing,1989,26(1):43-49.
[13]荆雷,马文君,常丹华.基于动态时间规整的手势加速度信号识别[J].传感器技术学报,2012,25(1):72-76.
[14]BELLMAN R. The theory of dynamic programming[J]. Bull.Amer.Math.Soc,1954,60(6):51-503.
[15]孙文,王刚,姚小强,等.临空高超声速飞行器目标特性分析[J].火力与指挥控制,2017,42(1):14-20.
[2]童奇,李建勋,童中翔,等.基于机动识别的空战意图威胁建模与仿真[J].现代防御技术,2014,42(4):174-184.
[3]王利芬,李莉,聂志强.超视距空战中目标机动意图评估[J].电光与控制,2012,19(12):68-71.
[4]谢川,倪世宏,章宗麟,等.一种基于知识的特技飞行动作快速识别方法[J].计算机工程,2004,30(12):116-118.
[5]尹文军,倪世宏.一种基于遗传算法的飞行动作识别方法[J].电脑开发与应用,2006,19(4):50-51.
[6]张军,陈付彬,付强,等.基于HMM的机动目标识别[J].国防科技大学学报,2003,25(2):51-55.
[7]宋元,王永春.海上防空作战态势估计理论及应用[M].北京:国防工业出版社,2014:100-126.
[8]ANDERSSON M.Classification of aerial missions using hidden markov models[C]//Proceedings of the 7th European Conference Symbolic and Quantitative Approaches to Reasoning with Uncertainty,2003:125-136.
[9]RABINER L R. A tutorial on hidden markov models and selected applications in speech recognition[J].IEEE Transactions on Acoustice,Speech and Signal Processing 1989,77(2):257-286.
[10]王芳,吴志泉,史红权. SVM在空中目标威胁值评估中的应用[J].火力与指挥控制,2017,42 (9):30-33.
[11]覃鸿,王守觉.多权值神经元网络仿生模式识别方法在低训练样本数量非特定人语音识别中与H MM及DTW的比较研究[J].电子学报,2005,33(5):957-960.
[12]SAKOE H,CHIBA S. Dynamic programming algorithm optimization for spoken word recognition[J]. IEEE Transactions on Acoustice,Speech and Signal Processing,1989,26(1):43-49.
[13]荆雷,马文君,常丹华.基于动态时间规整的手势加速度信号识别[J].传感器技术学报,2012,25(1):72-76.
[14]BELLMAN R. The theory of dynamic programming[J]. Bull.Amer.Math.Soc,1954,60(6):51-503.
[15]孙文,王刚,姚小强,等.临空高超声速飞行器目标特性分析[J].火力与指挥控制,2017,42(1):14-20.