Handwritten digit recognition based on ghost imaging with deep learning

doi:10.1088/1674-1056/abd2a5

中国物理B ›› 2021, Vol. 30 ›› Issue (5): 54201-054201.doi: 10.1088/1674-1056/abd2a5

Handwritten digit recognition based on ghost imaging with deep learning

Xing He(何行)¹, Sheng-Mei Zhao(赵生妹)^1,2,†, and Le Wang(王乐)²

1 Institute of Signal Processing and Transmission, Nanjing University of Posts and Telecommunications, Nanjing 210003, China;
2 Key Laboratory of Broadband Wireless Communication and Sensor Network Technology(Ministry of Education), Nanjing 210003, China

收稿日期:2020-08-10 修回日期:2020-11-17 接受日期:2020-12-11 出版日期:2021-05-14 发布日期:2021-05-14
通讯作者: Sheng-Mei Zhao E-mail:zhaosm@njupt.edu.cn
基金资助:
Project supported by the National Natural Science Foundation of China (Grant Nos. 61871234 and 11847062).

Handwritten digit recognition based on ghost imaging with deep learning

Xing He(何行)¹, Sheng-Mei Zhao(赵生妹)^1,2,†, and Le Wang(王乐)²

1 Institute of Signal Processing and Transmission, Nanjing University of Posts and Telecommunications, Nanjing 210003, China;
2 Key Laboratory of Broadband Wireless Communication and Sensor Network Technology(Ministry of Education), Nanjing 210003, China

Received:2020-08-10 Revised:2020-11-17 Accepted:2020-12-11 Online:2021-05-14 Published:2021-05-14
Contact: Sheng-Mei Zhao E-mail:zhaosm@njupt.edu.cn
Supported by:
Project supported by the National Natural Science Foundation of China (Grant Nos. 61871234 and 11847062).

摘要/Abstract

摘要： We present a ghost handwritten digit recognition method for the unknown handwritten digits based on ghost imaging (GI) with deep neural network, where a few detection signals from the bucket detector, generated by the cosine transform speckle, are used as the characteristic information and the input of the designed deep neural network (DNN), and the output of the DNN is the classification. The results show that the proposed scheme has a higher recognition accuracy (as high as 98% for the simulations, and 91% for the experiments) with a smaller sampling ratio (say 12.76%). With the increase of the sampling ratio, the recognition accuracy is enhanced. Compared with the traditional recognition scheme using the same DNN structure, the proposed scheme has slightly better performance with a lower complexity and non-locality property. The proposed scheme provides a promising way for remote sensing.

关键词: ghost imaging, handwritten digit recognition, ghost handwritten recognition, deep learning

Abstract: We present a ghost handwritten digit recognition method for the unknown handwritten digits based on ghost imaging (GI) with deep neural network, where a few detection signals from the bucket detector, generated by the cosine transform speckle, are used as the characteristic information and the input of the designed deep neural network (DNN), and the output of the DNN is the classification. The results show that the proposed scheme has a higher recognition accuracy (as high as 98% for the simulations, and 91% for the experiments) with a smaller sampling ratio (say 12.76%). With the increase of the sampling ratio, the recognition accuracy is enhanced. Compared with the traditional recognition scheme using the same DNN structure, the proposed scheme has slightly better performance with a lower complexity and non-locality property. The proposed scheme provides a promising way for remote sensing.

Key words: ghost imaging, handwritten digit recognition, ghost handwritten recognition, deep learning

中图分类号: (Pattern recognition)

42.30.Sy

42.30.-d (Imaging and optical processing)

Xing He(何行), Sheng-Mei Zhao(赵生妹), and Le Wang(王乐). Handwritten digit recognition based on ghost imaging with deep learning[J]. 中国物理B, 2021, 30(5): 54201-054201.

Xing He(何行), Sheng-Mei Zhao(赵生妹), and Le Wang(王乐). Handwritten digit recognition based on ghost imaging with deep learning[J]. Chin. Phys. B, 2021, 30(5): 54201-054201.

参考文献

[1] Cun Y L, Boser B, Denker J S, Henderson D and Jackel L D 1990 NIPS'89: Proceedings of the 2nd International Conference on Neural Information Processing Systems pp. 396-404
[2] Wang Y, Wang X and Liu W 2016 Inf. Sci. 351 67
[3] Goltsev A and Gritsenko V 2012 Neural Netw. 28 15
[4] Varatharajan R, Manogaran G and Priyan M K 2018 Multimed Tools Appl. 77 10195
[5] Kulkarni S R and Rajendran B 2018 Neural Netw. 103 118
[6] Li H X, Bai Y F, Shi X H, Nan S Q, Qu L J, Shen Q and Fu X Q 2017 Chin. Phys. B 26 104204
[7] Yin M Q, Wang L and Zhao S M 2019 Chin. Phys. B 28 094201
[8] Wang L, Zou L and Zhao S M 2018 Opt. Commun. 407 181
[9] Ren H D, Wang L and Zhao S M 2019 OSA Continuum 2 64
[10] Chen W and Chen X 2019 Appl. Phys. Lett. 104 251109
[11] Jiao S, Zhou C, Shi Y, Zou W and Li X 2019 Opt. Laser Technol. 109 370
[12] Zhang Y D and Zhao S M 2017 Chin. Phys. B 26 054205
[13] Zhao S, Wang L, Liang W, Cheng W and Gong L 2015 Opt. Commun. 353 90
[14] Strekalov D V, Sergienko A V, Klyshko D N and Shih Y H 1995 Phys. Rev. Lett. 74 3600
[15] Bennink R S, Bentley S J and Boyd R W 2002 Phys. Rev. Lett. 89 113601
[16] Valencia A, Scarcelli G, D'Angelo M and Shih Y 2005 Phys. Rev. Lett. 94 063601
[17] Shapiro J H 2008 Phys. Rev. A 78 061802
[18] Latorre-Carmona P, Traver V J, Sanchez J S and Tajahuerce E 2019 Image and Vision Computing 86 28
[19] Jiao S, Feng J, Gao Y, Lei T, Xie Z and Yuan X 2019 Opt. Lett. 44 5186
[20] Chen H, Shi J, Liu X, Niu Z and Zeng G 2018 Opt. Commun. 413 269
[21] Xu Y and Kelly K F 2019 arXiv:1901.09983 [cs.CV]
[22] Bacca J, Correa C V, Vargas E, Castillo S and Arguello H 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing (MLSP), October 13-16, 2019, Pittsburgh, PA, p. 1
[23] Zhong J, Zhang Z, Li X, Zheng S and Zheng G 2020 Opt. Express 9 28
[24] Ahmed N, Natarajan T, Rao and K R 1974 IEEE Trans. Comput. 23 90
[25] Kingma D P and Ba J L 2014 arXiv:1412.6980v8 [cs.LG]
[26] Wen Z, He B, Kotagiri R, Lu S and Shi J 2018 IEEE International Parallel and Distributed Processing Symposium (IPDPS), May 21-25, 2018, Vancouver, Canada, p. 234
[27] Dogru N and Subasi A 2018 15th Learning and Technology Conference, February 25-26, 2018, Jeddah, Saudi Arabia, p. 40
[28] Wang F, Wang H, Haichao W, Li G and Situ G 2019 Opt. Express 27 25560
[29] Devi R G and Sumanjani P 2015 IEEE International Conference on Engineering and Technology (ICETECH), March 20, 2015, Coimbatore, India, p. 95

Handwritten digit recognition based on ghost imaging with deep learning

Handwritten digit recognition based on ghost imaging with deep learning

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	Zhong-Yuan Liu(刘忠源), Shao-Ying Meng(孟少英), and Xi-Hao Chen(陈希浩)^†. A probability theory for filtered ghost imaging[J]. 中国物理B, 2023, 32(4): 44204-044204.
[2]	Zhao-Qi Liu(刘兆骐), Yan-Feng Bai(白艳锋), Xuan-Peng-Fan Zou(邹璇彭凡), Li-Yu Zhou(周立宇), Qin Fu(付芹), and Xi-Quan Fu(傅喜泉). Ghost imaging based on the control of light source bandwidth[J]. 中国物理B, 2023, 32(3): 34210-034210.
[3]	Xiao-Gang Wang(汪小刚) and Hao-Yu Wei(魏浩宇). Deep-learning-based cryptanalysis of two types of nonlinear optical cryptosystems[J]. 中国物理B, 2022, 31(9): 94202-094202.
[4]	Hui Guo(郭辉), Le Wang(王乐), and Sheng-Mei Zhao(赵生妹). Imaging a periodic moving/state-changed object with Hadamard-based computational ghost imaging[J]. 中国物理B, 2022, 31(8): 84201-084201.
[5]	Jin-Fen Liu(刘进芬), Le Wang(王乐), and Sheng-Mei Zhao(赵生妹). Orthogonal-triangular decomposition ghost imaging[J]. 中国物理B, 2022, 31(8): 84202-084202.
[6]	Xun Guo(郭寻), Hao Wang(王浩), Changkai Li(李长楷),Shijun Zhao(赵仕俊), Ke Jin(靳柯), and Jianming Xue(薛建明). Development of an electronic stopping power model based on deep learning and its application in ion range prediction[J]. 中国物理B, 2022, 31(7): 73402-073402.
[7]	Li-Jun Chang(常莉君), Yi-Fan Mo(莫一凡), Li-Ming Ling(凌黎明), and De-Lu Zeng(曾德炉). Data-driven parity-time-symmetric vector rogue wave solutions of multi-component nonlinear Schrödinger equation[J]. 中国物理B, 2022, 31(6): 60201-060201.
[8]	Haipeng Zhang(张海鹏), Ke Li(李可), Changzhe Zhao(赵昌哲), Jie Tang(汤杰), and Tiqiao Xiao(肖体乔). Efficient implementation of x-ray ghost imaging based on a modified compressive sensing algorithm[J]. 中国物理B, 2022, 31(6): 64202-064202.
[9]	Gaoyi Lei(雷高益), Chencheng Tang(唐陈成), and Yueyang Zhai(翟跃阳). Fringe removal algorithms for atomic absorption images: A survey[J]. 中国物理B, 2022, 31(5): 50313-050313.
[10]	Xun-Qin Huo(火勋琴), Wei-Feng Yang(杨玮枫), Wen-Hui Dong(董文卉), Fa-Cheng Jin(金发成), Xi-Wang Liu(刘希望), Hong-Dan Zhang(张宏丹), and Xiao-Hong Song(宋晓红). Review on typical applications and computational optimizations based on semiclassical methods in strong-field physics[J]. 中国物理B, 2022, 31(3): 33101-033101.
[11]	Qiwen Xu(徐启文), Zhu Zheng(郑铸), and Huabei Jiang(蒋华北). Deep learning for image reconstruction in thermoacoustic tomography[J]. 中国物理B, 2022, 31(2): 24302-024302.
[12]	Shao-Ying Meng(孟少英), Mei-Yi Chen(陈美伊), Jie Ji(季杰), Wei-Wei Shi(史伟伟), Qiang Fu(付强), Qian-Qian Bao(鲍倩倩), Xi-Hao Chen(陈希浩), and Ling-An Wu(吴令安). Iterative filtered ghost imaging[J]. 中国物理B, 2022, 31(2): 28702-028702.
[13]	Tian-Shou Liang(梁添寿), Peng-Peng Shi(时朋朋), San-Qing Su(苏三庆), and Zhi Zeng(曾志). Learning physical states of bulk crystalline materials from atomic trajectories in molecular dynamics simulation[J]. 中国物理B, 2022, 31(12): 126402-126402.
[14]	Chengwei Deng(邓成伟), Yunxin Tang(唐蕴芯), Jian Zhang(张建), Wenfei Li(李文飞), Jun Wang(王骏), and Wei Wang(王炜). RNAGCN: RNA tertiary structure assessment with a graph convolutional network[J]. 中国物理B, 2022, 31(11): 118702-118702.
[15]	Le Wang(王乐), Hui Guo(郭辉), and Shengmei Zhao(赵生妹). Full color ghost imaging by using both time and code division multiplexing technologies[J]. 中国物理B, 2022, 31(11): 114202-114202.