基于增量学习的合成孔径雷达目标识别算法
|
摘要
传统的合成孔径雷达(SAR)目标识别往往采用批量学习的方法,但是在现实应用中,系统的训练数据并不能一次性全部获得,当有新的训练样本到来时,采用批量学习方法需要重新训练整个系统。为解决这个问题,将增量学习算法——正则在线序贯式极限学习机(ROSELM)应用到SAR目标识别中,并且采用粒子群算法优化ROSELM的初始权值以提高其稳定性和识别率。实验结果表明,该算法在新的SAR目标样本获得时只需要通过更新输出权值即可完成系统的更新,无需重新训练,且速度极快、识别率高,可以作为SAR目标识别系统在线更新的良好选择。
Target recognition of conventional Synthetic Aperture Radar( SAR) usually adopts the batch learning method. However in practical application the training data of the system can't be obtained all at one time. When new training sample arrives the whole system needs to be retrained when using the method of batch learning. In order to solve this problem ROSELM an incremental learning algorithm is applied to SAR target recognition and Particle Swarm Optimization( PSO) algorithm is used to optimize the initial weight of ROSELM to improve its stability and recognition rate. The experimental results show that: 1) When new SAR target samples are obtained the system updating can be implemented simply by updating the output weights without re-training; 2) The algorithm is very fast and has a high recognition rate which is a good choice for online updating of SAR target recognition system.
Target recognition of conventional Synthetic Aperture Radar( SAR) usually adopts the batch learning method. However in practical application the training data of the system can't be obtained all at one time. When new training sample arrives the whole system needs to be retrained when using the method of batch learning. In order to solve this problem ROSELM an incremental learning algorithm is applied to SAR target recognition and Particle Swarm Optimization( PSO) algorithm is used to optimize the initial weight of ROSELM to improve its stability and recognition rate. The experimental results show that: 1) When new SAR target samples are obtained the system updating can be implemented simply by updating the output weights without re-training; 2) The algorithm is very fast and has a high recognition rate which is a good choice for online updating of SAR target recognition system.
引文
[1]文贡坚,朱国强,殷红成,等.基于三维电磁散射参数化模型的SAR目标识别方法[J].雷达学报,2017,6(2):115-135.
[2]雷磊,杨秋,李开明.基于多尺度稀疏字典的SAR图像目标识别方法[J].火力与指挥控制,2017,42(4):10-13.
[3]耿杰,范剑超,初佳兰,等.基于深度协同稀疏编码网络的海洋浮筏SAR图像目标识别[J].自动化学报,2016,42(4):593-604.
[4]丁军,刘宏伟,陈渤,等.相似性约束的深度置信网络在SAR图像目标识别的应用[J].电子与信息学报,2016,38(1):97-103.
[5]丁军,刘宏伟,王英华.基于非负稀疏表示的SAR图像目标识别方法[J].电子与信息学报,2014,36(9):2194-2200.
[6] LIANG N YHUANG G B,SARATCHANDRAN P,et al.A fast and accurate online sequential learning algorithm for feed forward networks[J]. IEEE Transactions on Networks200617(6):1411-1423.
[7] HUANG G B,ZHU Q YSIEW C K. Extreme learning machine:theory and applications[J]. Neurocomputing200670(1-3):489-501.
[8]许学斌,管晓宏,邓万宇,等.基于多分类投影极速学习机的快速人脸识别方法[J].微电子学与计算机,2012,29(7):13-17.
[9] ZONG W WHUANG G B. Face recognition based on extreme learning machine[J]. Neurocomputing201174(16):2541-2551.
[10]易子麟,尹东,胡安洲,等.基于非局部均值滤波的SAR图像去噪[J].电子与信息学报,2012,34(4):950-955.
[11] BUADES ACOLL BMOREL J M. A non-local algorithm for image denoising[C]//CVPR IEEE Computer Society Conference on Computer Vision and Pattern RecognitionIEEE2005:60-65.
[12]王通,高宪文,蔺雪,等. SWE-IPCA方法在传感器故障诊断中的应用[J].仪器仪表学报,2013,34(8):1841-1846.
[2]雷磊,杨秋,李开明.基于多尺度稀疏字典的SAR图像目标识别方法[J].火力与指挥控制,2017,42(4):10-13.
[3]耿杰,范剑超,初佳兰,等.基于深度协同稀疏编码网络的海洋浮筏SAR图像目标识别[J].自动化学报,2016,42(4):593-604.
[4]丁军,刘宏伟,陈渤,等.相似性约束的深度置信网络在SAR图像目标识别的应用[J].电子与信息学报,2016,38(1):97-103.
[5]丁军,刘宏伟,王英华.基于非负稀疏表示的SAR图像目标识别方法[J].电子与信息学报,2014,36(9):2194-2200.
[6] LIANG N YHUANG G B,SARATCHANDRAN P,et al.A fast and accurate online sequential learning algorithm for feed forward networks[J]. IEEE Transactions on Networks200617(6):1411-1423.
[7] HUANG G B,ZHU Q YSIEW C K. Extreme learning machine:theory and applications[J]. Neurocomputing200670(1-3):489-501.
[8]许学斌,管晓宏,邓万宇,等.基于多分类投影极速学习机的快速人脸识别方法[J].微电子学与计算机,2012,29(7):13-17.
[9] ZONG W WHUANG G B. Face recognition based on extreme learning machine[J]. Neurocomputing201174(16):2541-2551.
[10]易子麟,尹东,胡安洲,等.基于非局部均值滤波的SAR图像去噪[J].电子与信息学报,2012,34(4):950-955.
[11] BUADES ACOLL BMOREL J M. A non-local algorithm for image denoising[C]//CVPR IEEE Computer Society Conference on Computer Vision and Pattern RecognitionIEEE2005:60-65.
[12]王通,高宪文,蔺雪,等. SWE-IPCA方法在传感器故障诊断中的应用[J].仪器仪表学报,2013,34(8):1841-1846.