臺灣博碩士論文加值系統

工業技術從原先著重於專業相關技術的開發與突破，轉型成重視其上下游產業鏈的一條龍客製化服務；工業技術的成就在實際應用上需要考慮到使用者的操作難易度以及適性貼合度，適應各類環境來改變工業技術的柔軟度，比起傳統專注於硬實力的提升，如今的自動化工業更講求客製化的服務，產出的物品或技術是否能契合需求，因此發展出物聯網(Internet of Things, IoT)系統；物聯網系統將硬體裝置結合網際網路，在不需要額外人力的情況下，便能自主收集與統合整體物聯網內部設施運行的數據統合，透過網際網路傳輸至雲端或者是專用伺服器進行數位資料的儲存與大數據計算分析，實現更為精確行為預測，進一步改善產業價值。
本研究將比較多種例如全卷積神經網路演算法如Fully Convolutional Networks (FCN)、High-Resolution Net (HRNet)等方法，以減少硬體加速所需要消耗的功率，並應用於工業物聯網；本研究將神經網路配合物聯網專注於即時分享感測器所回傳之資訊，完整建立可移動式平台的神經網路於生產場域的應用。

In this paper, a new algorithm combines a Full Convolutional Network (FCN) to identify a rough floor area, and then using Canny Edge Detection to identify a rough floor area, and intersect the two results. In order to achieve real-time performance, the proposed algorithm has been implemented on an NVIDIA Jetson TX2 embedded platform. In experimental results, a public MIT scene dataset and indoor database were used to verify detection accuracy. Compared to recent works, the proposed FCN based algorithm accuracy can reach up to 97.1% on average without the assistance of any other physical sensors, such as RGB-D or laser ranger sensors.

摘要............................i
Abstract........................ii
誌謝............................iii
目錄............................iv
表目錄..........................v
圖目錄..........................vi
第一章 緒論 1
1.1 研究背景與動機..........1
1.2 研究方法................3
1.3 論文組織................4
第二章 背景知識與相關研究......5
2.1 卷積神經網路............5
2.2 全卷積神經網路(Fully Convolutional Network, FCN)........7
2.3 語意分割網路(Semantic Segmentation Networks, Seg-Net)...9
2.4 高解析網路(High-Resolution Networks, HR-Net)............10
2.5 神經網路量化(Neural Network Quantization)...............11
第三章 系統架構與硬體設計..............14
3.1 頭戴式AR虛擬擴增實境裝置........14
3.2 OPC 統一架構....................15
3.3 JETSON XAVIER NX................18
3.4 自動導引運輸車..................19
3.5 ROS機器人作業系統...............20
第四章 實驗流程與結果..................22
4.1 機器人系統流程..................22
4.2 OPC Server 資料連結.............23
4.3 HoloLens擴增實境................24
4.4 硬體整合........................25
4.5 東元無人搬運機器人..............26
4.6 網路測試結果....................27
4.7 整合結果........................30
第五章 結論............................34
參考文獻................................35
Extended Abstract......................37

[1]P. Ongsulee, "Artificial intelligence, machine learning and deep learning," in International Conference on ICT and Knowledge Engineering, 2017, pp. 1-6.
[2]R. Brandon, A. Robert, W. Paul, W. Gu-Yeon, B. David, and M. Margaret, Deep Learning for Computer Architects. 2017.
[3]EFFRA. (2019 ). Factories of the Future. Available: https://www.effra.eu/factories-future
[4]A.-C. H. a. J. Krogstie, "Mobile augmented reality for cultural heritage: A technology acceptance study," presented at the 2012 IEEE International Symposium on Mixed and Augmented Reality (ISMAR), 2012.
[5]S. K. a. P. K. M. Lorenz, "Industrial Augmented Reality: Requirements for an Augmented Reality Maintenance Worker Support System," presented at the IEEE International Symposium on Mixed and Augmented Reality Workshops (ISMARW), 2018.
[6]D. A. B. a. R. P. McMahan, "Virtual Reality: How Much Immersion Is Enough?," Computer, vol. 40, no. 7, pp. 36-43, 2007.
[7]B. O. C. Bichlmeier, S. M. Heining, A. Ahmadi and N., "Stepping into the operating theater: ARAV — Augmented Reality Aided Vertebroplasty," presented at the 7th IEEE/ACM International Symposium on Mixed and Augmented Reality, 2008.
[8]M. S. a. L. Teixeira, "Developing an eXtended Reality platform for Immersive and Interactive Experiences for Cultural Heritage: Serralves Museum and Coa Archeologic Park," presented at the 2020 IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct), 2020.
[9]A. C. M. E. C. Santos, T. Taketomi, G. Yamamoto, J. Miyazaki and H. Kato, "Augmented Reality Learning Experiences: Survey of Prototype Design and Evaluation," IEEE Transactions on Learning Technologies, vol. 7, pp. 38-56.
[10]D. Miyashita, E. H. L. and, and B. Murmann, "Convolutional Neural Networks using Logarithmic Data Representation," CoRR, vol. abs/1603.01025, 2016.
[11]M. Bettoni, G. Urgese, Y. Kobayashi, E. Macii, and A. Acquaviva, "A Convolutional Neural Network Fully Implemented on FPGA for Embedded Platforms," in 2017 New Generation of CAS (NGCAS), pp. 49-52.
[12]E. Shelhamer, J. Long, and T. Darrell, "Fully Convolutional Networks for Semantic Segmentation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 4, pp. 640-651, 2017.
[13]V. Badrinarayanan, A. Kendall, and R. Cipolla, "SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 12, pp. 2481-2495, 2017.
[14]V. Badrinarayanan, A. Handa, and R. Cipolla, "SegNet: A Deep Convolutional Encoder-Decoder Architecture for Robust Semantic Pixel-Wise Labelling," CoRR, vol. abs/1505.07293, 2015.
[15]J. Wang et al., "Deep High-Resolution Representation Learning for Visual Recognition," IEEE transactions on pattern analysis and machine intelligence, 2020.
[16]M. C. a. Y. Bengio, "BinaryNet: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1," Computing of Research Respoitory, vol. abs/1602.02830, 2016.
[17]I. H. and, M. C. and, D. S. and, R. E.-Y. and, and Y. Bengio, "Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations," CoRR, vol. abs/1609.07061, 2016.
[18]OPC Unified Architecture Part 1: Overview and Concepts, 2017.
[19]J. M. O'Kane, A Gentle Introduction to ROS. CreateSpace Independent Publishing Platform, 2013.
[20]M. Quigley, B. Gerkey, and W. D. Smart, Programming Robots with ROS. 2015.