該研究提出了一種低成本有效的人工智能（AI）算法集成系統，用於商務辦公中經常使用的移動設備的通信。移動設備包括安卓、基於 ROS 的設備和機器人。用於通信的工具包括藍牙和串列通訊。在研究中，使用名為基於區域的卷積神經網絡 (R-CNN) 的 AI 算法來偵測對象。該研究的目的是將所有應用程序集成在一起，以提高數據通信和監控的效率。

The study proposed a low-cost and effective artificial intelligence (AI) algorithm integrated system for communication of mobile devices that are frequently used in a business office. The mobile devices include android, ROS based devices, and robots. The tools used for communication include Bluetooth and serial port. In the study, an AI algorithm named region-based Convolutional Neural Networks (R-CNN) was used to detect object. The objective of the study is to integrate all applications together in order to improve the efficiency of data communication and monitoring.

Table of contents	III
	Chapter 1 Introduction	1
1.1	Background	1
1.2	Objectives	1
1.3	Scopes	1
	Chapter 2 Literature Reviews	2
	Chapter 3 Related theories	6
3.1	Tensor flow	6
3.2	Tensor RT	7
3.3	Convolutional Neural Networks (CNN)	7
3.4	Regional-based Convolutional Neural Networks (R-CNN)	7
3.5	You Only Look Once (YOLO)	8
3.6	Fovea Box	8
3.7	Bluetooth	9
3.7.1 L2CAP	10
3.7.2 RFCOMM port	10
3.8	Overview of the thesis	11
	Chapter 4 Robot Model and Algorithm Design	12
4.1	Hardware Architecture	12
4.1.1 4Robot Model	12
4.1.2 Jetson and board	21
4.2	Algorithm	24
4.2.1 AI algorithms	24
4.2.2 Distance calculation	25
4.3	Mobile Application	25
	Chapter 5 Experiment	27
5.1	Mobile application	27
5.2	Bluetooth server	31
5.3	Object detection	32
5.4	Video Transfer	33
	Chapter 6 Conclusion and Discussion	35
6.1	Conclusion	35
6.2	Discussion	35
	Appendix A Internship in Aetina company	36
References	38

List of Figures
Figure 1 System Overview Architecture	1
Figure 2 R-CNN pipeline with ROI Pool and ROI Align [22]	7
Figure 3 YOLO Network [9]	8
Figure 4 Bluetooth Architecture [16]	9
Figure 5 RFCOMM Structure	10
Figure 6 Overview Structure Architecture	11
Figure 7 Turtlebot3 Burger	13
Figure 8 OpenCR Board	15
Figure 9 OpenCR structure	16
Figure 10 Jetson Xavier NX	21
Figure 11 System Flow	26
Figure 12 Application’s page	27
Figure 13 Bluetooth Device	28
Figure 14 Report file	29
Figure 15 Jetson Data on Mobile Application	29
Figure 16 Internet connection	30
Figure 17 Control view on Mobile Application	31

[1]	T. Kong, F. Sun, H. Liu, Y. Jiang, L. Li, and J. Shi, "Foveabox: Beyound anchor-based object detection," IEEE Transactions on Image Processing, vol. 29, pp. 7389-7398, 2020.
[2]	H. Zhang, K. Watanabe, K. Motegi, and Y. Shiraishi, "ROS Based Framework for Autonomous Driving of AGVs," Proceedings of ICMEMIS, pp. 4-6, 2019.
[3]	H. Hu, J. Gu, Z. Zhang, J. Dai, and Y. Wei, "Relation networks for object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3588-3597.
[4]	S. Ren, K. He, R. Girshick, and J. Sun, "Faster r-cnn: Towards real-time object detection with region proposal networks," Advances in neural information processing systems, vol. 28, pp. 91-99, 2015.
[5]	S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: towards real-time object detection with region proposal networks," IEEE transactions on pattern analysis and machine intelligence, vol. 39, no. 6, pp. 1137-1149, 2016.
[6]	R. Girshick, "Fast r-cnn," in Proceedings of the IEEE international conference on computer vision, 2015, pp. 1440-1448.
[7]	J. Dai, Y. Li, K. He, and J. Sun, "R-fcn: Object detection via region-based fully convolutional networks," in Advances in neural information processing systems, 2016, pp. 379-387.
[8]	Z. Li, C. Peng, G. Yu, X. Zhang, Y. Deng, and J. Sun, "Detnet: A backbone network for object detection," arXiv preprint arXiv:1804.06215, 2018.
[9]	J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You only look once: Unified, real-time object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 779-788.
[10]	R. Huang, J. Pedoeem, and C. Chen, "YOLO-LITE: a real-time object detection algorithm optimized for non-GPU computers," in 2018 IEEE International Conference on Big Data (Big Data), 2018: IEEE, pp. 2503-2510.
[11]	C. Szegedy, A. Toshev, and D. Erhan, "Deep neural networks for object detection," 2013.
[12]	A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," Advances in neural information processing systems, vol. 25, pp. 1097-1105, 2012.
[13]	S. Banerjee, D. Mondal, S. Das, and R. B. Guin, "Real-time video streaming over Bluetooth network between two mobile nodes," International Journal of Computer Science (IJCSI), vol. 7, no. 3, pp. 37-39, 2010.
[14]	D. Catania and S. Zammit, "Video streaming over Bluetooth," 2008.
[15]	X. Wang, "Video Streaming over Bluetooth," Institute for Infocoram Research (12R), School of Computing, National University of Singapore, 2004.
[16]	S. Gupta, S. K. Singh, and R. Jain, "Analysis and optimisation of various transmission issues in video streaming over Bluetooth," International Journal of Computer Applications, vol. 11, no. 7, pp. 44-48, 2010.
[17]	M. Abadi et al., "Tensorflow: A system for large-scale machine learning," in 12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16), 2016, pp. 265-283.
[18]	M. Abadi et al., "Tensorflow: Large-scale machine learning on heterogeneous distributed systems," arXiv preprint arXiv:1603.04467, 2016.
[19]	Object_detection. Website github.com: https://github.com/tensorflow/models/tree/master/res earch/object_detection. sited in May 2021.
[20]	Cudnn. nvidia.com: https://developer.nvidia.com/cudnn. sited in May 2021.
[21]	H. Vanholder, "Efficient inference with tensorrt," ed: ed, 2016.
[22]	L. Yang, Q. Song, Z. Wang, and M. Jiang, "Parsing r-cnn for instance-level human analysis," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 364-373.
[23]	RFCOMM, Website: https://www.programmersought.com/article/9546874968/, Sited in May2021.
[24]	TurtleBot feature. Website: https://emanual.robotis.com/docs/en/platform/turtlebot3/feature s/. sited in May 2021.
[25]	COCO test-dev standard, Website: https://paperswithcode.com/sota/object-detection-on-coco, Sited in May 2021.
[26]	A. Sharif Razavian, H. Azizpour, J. Sullivan, and S. Carlsson, "CNN features off-the-shelf: an astounding baseline for recognition," in Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 2014, pp. 806-813.