Research on UAV target recognition algorithm based on transfer learning SAE

Xie Bing; Duan Zhemin; Zheng Bin; Yin Yunhua

doi:10.3788/IRLA201847.0626001

UAV in complex battlefield environment, because the two sides of the UAV shape, color and other characteristics are more similar, how to identity enemy UAV accurately is the key to realize the autonomous navigation and combat mission execution. However, due to changes in the speed, shape, size, attitude of enemy UAV and the impact of meteorological and environmental factors, they can not be accurately identified and classified. Aiming at this problem, a kind of sparse auto-encoder(SAE) based on the transfer learning was proposed, and the target objects in the multi-frame aerial images were identified. The algorithm firstly used SAE to study the unsupervised learning of a large number of unmarked samples in the data concentration of source domain, and obtained its local characteristics. Then, the global feature response of the aerial image in the target domain was extracted by the convolution neural network (CNN) algorithm. Finally, the different categories of target objects were identified and classified by the Softmax regression model. The experimental results show that new algorithm proposed in this paper for multiple target objects in aerial multi-frame images is better than more traditional non-transfer learning SAE algorithm, and underlying visual feature recognition transfer learning algorithm, which has higher recognition rate.

Research on UAV target recognition algorithm based on transfer learning SAE

1. School of Electronics and Information,Northwestern Polytechnical University,Xi'an 710072,China;

2. School of Electrical and Control Engineering,North University of China,Taiyuan 030051,China

Received Date: 2018-01-02

Rev Recd Date: 2018-02-18

Publish Date: 2018-06-25

Key words:

autonomous navigation of UAV /
classification and identification of target objects /
sparse auto-encoder /
convolutional neural network /
transfer learning

Abstract: UAV in complex battlefield environment, because the two sides of the UAV shape, color and other characteristics are more similar, how to identity enemy UAV accurately is the key to realize the autonomous navigation and combat mission execution. However, due to changes in the speed, shape, size, attitude of enemy UAV and the impact of meteorological and environmental factors, they can not be accurately identified and classified. Aiming at this problem, a kind of sparse auto-encoder(SAE) based on the transfer learning was proposed, and the target objects in the multi-frame aerial images were identified. The algorithm firstly used SAE to study the unsupervised learning of a large number of unmarked samples in the data concentration of source domain, and obtained its local characteristics. Then, the global feature response of the aerial image in the target domain was extracted by the convolution neural network (CNN) algorithm. Finally, the different categories of target objects were identified and classified by the Softmax regression model. The experimental results show that new algorithm proposed in this paper for multiple target objects in aerial multi-frame images is better than more traditional non-transfer learning SAE algorithm, and underlying visual feature recognition transfer learning algorithm, which has higher recognition rate.

HTML

Reference (16)

[1]	Li Min, Zhou Zhenhua, Zhang Guilin. Image metric in performance evaluation of automatic target recognition algorithm[J]. Infrared and Laser Engineering, 2007, 36(3):412-416. (in Chinese)
[2]	Jia Ping, Xu Ning, Zhang Ye. Automatic target recognition based on local feature extraction[J]. Optics and Precision Engineering, 2013, 21(7):1898-1905. (in Chinese)
[3]	Xu Fang, Liu Jinghong, Zeng Dongdong, et al. Unsupervised ship detection and recognition based on Visual salience[J]. Optics and Precision Engineering, 2017, 25(5):1300-1311. (in Chinese)
[4]	Jiang Shan, Zhang Rui, Han Guangliang, et al. Multi-feature fusion moving target tracking in gray image with complex background[J]. Chinese Optics, 2016, 9(3):320-328. (in Chinese)
[5]	Lecun Y, Bengio Y, Hinton G. Deep learning[J]. Nature, 2015, 521(7553):436-444.
[6]	Chen T, Borth D, Darrell T, et al. Deepsentibank:visual sentiment concept classification with deep convolutional neural networks[J]. Computer Science, 2014, 65(5):1087-1095.
[7]	You Q, Luo J, Jin H, et al. Robust image sentiment analysis using progressively trained and domain transferred deep networks[C]//29th AAAI Conference on Artificial Intelligence (AAAI), 2015:381-388.
[8]	Zhang F, Du B, and Zhang L. Saliency-guided unsupervised feature learning for scene classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(4):2175-2184.
[9]	Zhuang Fuzhen, Luo Ping, He Qing, et al. Progress in transfer learning[J]. Journal of Software, 2015, 261:26-39. (in Chinese)
[10]	Song Jianzhong. The development trend of intelligent image processing[J]. Chinese Optics, 2011, 4(5):431-440. (in Chinese)
[11]	Coates A, Lee H, Ng A Y. An analysis of single-layer networks in unsupervised feature learning[C]//14th International Conference on Artificial Intelligence and Statistics, 2011:215-223.
[12]	Eng J, Zhang Z, Eyben F, et al. Autoencoder-based unsupervised domain adaptation for speech emotion recognition[J]. IEEE Signal Processing Letters, 2014, 21(9):1068-1072.
[13]	Zhou J T, Pan S J, Tsang I W, et al. Hybrid heterogeneous transfer learning through deep learning[C]//28th AAAI Conference on Artificial Intelligence(AAAI), 2014:2213-2219.
[14]	Guo Yingkai, Yang Jie, Lu Zhenggang. Unsupervised texture segmentation based on gaussian markov random field and neural network[J]. Infrared and Laser Engineering, 2000, 29:225-229. (in Chinese)
[15]	Vedaldi A, Lenc K. MatConvNet:convolutional neural networks for matlab[C]//23rd ACM International Conference on Multimedia, 2015:689-692.
[16]	Luo Haibo, Xu Lingyun, Hui Bin, et al. Present situation and prospect of target tracking methods based on deep learning[J]. Infrared and Laser Engineering, 2017, 46(5):0502002. (in Chinese)

Research on UAV target recognition algorithm based on transfer learning SAE

doi: 10.3788/IRLA201847.0626001

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views