[1] [1] Sonka M, Václav Hlavác, Boyle R. Image Processing, Analysis Machine Vision[M]. New Yk: Cengage Learning, 2014.

[2] [2] Ishikawa M, Ogawa K, Komuro T, et al. A CMOS vision chip with SIMD processing element array f 1 ms image processing[C]IEEE Int SolidState Circuits Conf, 1999: 206207.

[3] W Miao, Q Lin, W Zhang. A programmable SIMD vision chip for real-time vision applications. IEEE Journal of Solid-State Circuits, 43, 1470-1479(2008).

[4] W Zhang, Q Fu, N Wu. A programmable vision chip based on multiple levels of parallel processors. IEEE Journal of Solid-State Circuits, 46, 2132-2147(2011).

[5] [5] Shi C, Yang J, Han Y, et al. A 1000 fps vision chip based on a dynamically reconfigurable hybrid architecture comprising a PE array selfganizing map neural wk[C]2014 IEEE International SolidState Circuits Conference Digest of Technical Papers (ISSCC), 2014: 128129.

[6] [6] Yamazaki T, Katayama H, Uehara S, et al. 4.9 A 1 ms highspeed vision chip with 3Dstacked 140GOPS columnparallel PEs f spatiotempal image processing[C]2017 IEEE International Solid State Circuits Conference (ISSCC), 2017: 8283.

[7] J Yang, Y Yang, Z Chen. A heterogeneous parallel processor for high-speed vision chip. IEEE Transactions on Circuits and Systems for Video Technology, 28, 746-758(2018).

[8] [8] Krizhevsky A, Sutskever I, Hinton G E. Image classification with deep convolutional neural wks[C]Advances in Neural Infmation Processing Systems. 2012: 10971105.

[9] V Sze, Y Chen, T Yang. Efficient processing of deep neural networks: A tutorial and survey. Proceedings of the IEEE, 105, 2295-2329(2017).

[10] T Chen, Z Du, N Sun. Diannao: A small-footprint high-throughput accelerator for ubiquitous machine-learning. ACM Sigplan Notices, 49, 269-284(2014).

[11] [11] Liu D, Chen T, Liu S, et al. Pudiannao: A polyvalent machine learning accelerat[C]ACM SIGARCH Computer Architecture News. ACM, 2015, 43(1): 369381.

[12] [12] Millet L, Chevobbe S, riamisaina C, et al. A 5500FPS 85GOPSW 3D stacked BSI vision chip based on parallel infocalplane acquisition processing[C]2018 IEEE Symposium on VLSI Circuits. IEEE, 2018: 245246.

[13] [13] Chen Y, Krishna T, Emer J, et al. 14.5 Eyeriss: An energyefficient reconfigurable accelerat f deep convolutional neural wks[C]2016 IEEE International SolidState Circuits Conference (ISSCC). IEEE, 2016: 262263.

[14] J Jo, S Cha, D Rho. DSIP: A scalable inference accelerator for convolutional neural networks. IEEE Journal of Solid-State Circuits, 53, 605-618(2017).