農業在人類文明發展上扮演著極為重要的角色，生產力的提升讓人類有餘力去發展創新科技。但現代中，雖有近代科技輔助，像是機械化農機具，基因改造科技和殺蟲劑等幫助。人們仍然面對諸如農場安全、人口劇增、作物疾病、牲畜疾病和動物異常行為等問題。因此增加食物產量及降低疾病帶來的損失是目前在農業和畜牧業的一個重要任務。前人的研究中使用照片訓練集來訓練AI影像辨識來達成對作物的相關分析研究。在畜牧業情況就變得比較複雜，雖然有一些動物相關的AI分析，例如動物品種辨識，也可以用照片來做訓練，但如果是跟動物行為就比較難用照片當作訓練集。因為動物的行為是一連串連續的動作，難以用單一照片來判斷行為的區別。本論文中設計並實作出一套基於雲端服務的全自動錄影串流系統，來解決收集影像資料的問題。該系統可以預約錄影時間，
讓攝影機在特定時間收集資料，也可以註冊事件觸發，讓攝影機能在特定重要事件觸發時開啟錄影功能記錄當下狀況，並也可以手動開啟錄影功能。最後我們使用AWS雲端整合服務，解決儲存空間限制、降低維護系統及開發新功能的難度，並可以同時管理全台灣多個場域的攝影機。

Agriculture plays a key role in human history. Increasing food productivity allows people to develop technology and civilization. Although benefited from modern technology, such as agricultural machinery and farming methods, genetic technology, and techniques for achieving economies of scale in production. We still face some problems, for example, farm security, population increase, plant diseases, cattle diseases, and abnormal behavior. Increasing productivity and decreasing disease damage of food resources in crops and stock are crucial missions for us. Previous researchers have used AI technology to do image recognition and classification for crop-related research by collecting photos as training data. But if it came to stock farming, this might become an issue. Although some analysis about an animal can be done by photo, abnormal behavior is hard to analyze because continuous behavior is the range of actions and mannerisms. In other words, it is hard to be recognized by a single photo. Video training data is the kind of thing we need. In this thesis, we propose a recording streaming system that is efficient, automatic, and cloud-based. It can not only record manually but also preset time scheduling for specific timing to record and register events to trigger record processes when something critical happened. Also, we use AWS as our cloud service which improves storage limit, the difficulty of maintaining the system and developing new features, etc. At last, we can manage all of the cameras in experimental fields located all over Taiwan.

1. Agriculture, https://en.wikipedia.org/wiki/Agriculture, accessed: 2022-10-2.
2. Diamond, J. (2002). Evolution, consequences and future of plant and animal domestica- tion. Nature, 418(6898), 700-707, 2002.
3. World population,https://www.worldometers.info/world-population/,accessed:2022-10- 2.
4. Borlaug NE (2000) The green revolution revisited and the road ahead. In: Special 30th Anniversary Lecture. The Norwegian Nobel Institute, Oslo, Norway, 2000.
5. Livestock, https://en.wikipedia.org/wiki/Livestock, accessed: 2022-10-2.
6. Suryawanshi, K. R., Redpath, S. M., Bhatnagar, Y. V., Ramakrishnan, U., Chaturvedi, V., Smout, S. C., et al. (2017). Impact of wild prey availability on livestock predation by snow leopards. R. Soc. Open Sci. 4:170026, 2017.
7. Bourne Jr, J. K. (2015). The end of plenty: The race to feed a crowded world. WW Norton & Company, 2015.
8. FAO, https://www.fao.org/home/en, accessed: 2022-10-2.
9. Hsu, C. W., Huang, Y. H., & Huang, N. F. (2022, January). Real-time Dragonfruit's Ripeness Classification System with Edge Computing Based on Convolution Neural Net- work. In 2022 International Conference on Information Networking (ICOIN) (pp. 177- 182). IEEE, 2022.
10. Omar Anas, Youssef Wageeh, Hussam El-Din Mohamed, Ali Fadl, Noha ElMasry, Ay- man Nabil, Ayman Atia, Detecting Abnormal Fish Behavior Using Motion Trajectories In Ubiquitous Environments, Procedia Computer Science, Volume 175, Pages 141-148, 2020.
11. Jingqiu, G., Zhihai, W., Ronghua, G., & Huarui, W. (2017). Cow behavior recognition based on image analysis and activities. International Journal of Agricultural and Biological Engineering, 10(3), 165-174, 2017.
12. Leroy,T.,Vranken,E.,VanBrecht,A.,Struelens,E.,Sonck,B.,&Berckmans,D.(2006). A computer vision method for on-line behavioral quantification of individually caged poultry. Transactions of the ASABE, 49(3), 795-802, 2006.
13. Smart Farming Platform, https://nthu-smart-farming.kits.tw/, accessed: 2022-10-2.
14. AWS cloud service, https://aws.amazon.com/tw/, accessed: 2022-10-2.
15. ’Whatisinternetofthings(IoT)?’,https://www.techtarget.com/iotagenda/definition/Internet- of-Things-IoT, accessed: 2022-10-2.
16. Atlam, H. F., & Wills, G. B. (2019). Intersections between IoT and distributed ledger. In Advances in Computers (Vol. 115, pp. 73-113). Elsevier, 2019.
17. InfoWorld Media Group, Inc. (2 March 1998). InfoWorld. InfoWorld Media Group, Inc. p. 18. ISSN 0199-6649, 1998.
18. Rafael Osso (1999). Handbook of Emerging Communications Technologies: The Next Decade. CRC Press. p. 42. ISBN 978-1-4200-4962-6, 1999.
19. Macromedia, https://en.wikipedia.org/wiki/Macromedia, accessed: 2022-10-2.
20. RTMP introduction, https://www.wowza.com/blog/rtmp-streaming-real-time-messaging- protocol, accessed: 2022-10-2.
21. MQTT, https://mqtt.org/, accessed: 2022-10-2.
22. Pub-sub pattern, https://en.wikipedia.org/wiki/Publish-subscribe_pattern, accessed: 2022-10-2.
23. Naik, N. (2017, October). Choice of effective messaging protocols for IoT systems: MQTT, CoAP, AMQP and HTTP. In 2017 IEEE international systems engineering sym- posium (ISSE) (pp. 1-7). IEEE, 2017.
24. Ng, X. L., Ong, K. E., Zheng, Q., Ni, Y., Yeo, S. Y., & Liu, J. (2022). Animal Kingdom: A Large and Diverse Dataset for Animal Behavior Understanding. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 19023-19034), 2022.
25. Lamdouar, H., Yang, C., Xie, W., & Zisserman, A. (2020). Betrayed by motion: Camou- flaged object discovery via motion segmentation. In Proceedings of the Asian Conference on Computer Vision, 2020.
26. Fang, C., Zhang, T., Zheng, H., Huang, J., & Cuan, K. (2021). Pose estimation and be- havior classification of broiler chickens based on deep neural networks. Computers and Electronics in Agriculture, 180, 105863, 2021.
27. de Lima Weber, F., de Moraes Weber, V. A., Menezes, G. V., Junior, A. D. S. O., Alves, D. A., de Oliveira, M. V. M., ... & de Abreu, U. G. P. (2020). Recognition of Pantaneira cattle breed using computer vision and convolutional neural networks. Computers and Electronics in Agriculture, 175, 105548, 2020.
28. UAVs introduction, https://en.wikipedia.org/wiki/Unmanned_aerial_vehicle, accessed: 2022-10-2.
29. UAV, https://www.vicorpower.com/zh-cn/resource-library/articles/modular-power- delivery-networks-for-uavs, accessed: 2022-10-2.
30. CCTV camera, https://www.clearway.co.uk/news/the-different-types-of-cctv-camera- explained/, accessed: 2022-10-2.
31. CCTV introduction, https://www.linkedin.com/pulse/introduction-cctv-sean-shaheer-shaaban/, accessed: 2022-10-2.
32. GenBest Technology digital video recording system, http://www.genbest.com.tw/genbest/dvr/ec.htm, accessed: 2022-10-2.
33. Luda farm, https://www.luda.farm/farm-video-surveillance/, accessed: 2022-10-2.
34. Chen, G., Han, T. X., He, Z., Kays, R., & Forrester, T. (2014, October). Deep convolu- tional neural network based species recognition for wild animal monitoring. In 2014 IEEE international conference on image processing (ICIP) (pp. 858-862). IEEE, 2014.
35. AWS S3, https://aws.amazon.com/tw/s3/, accessed: 2022-10-2.
36. AWS lambda, https://aws.amazon.com/tw/lambda/, accessed: 2022-10-2.
37. AWS greengrass, https://aws.amazon.com/tw/greengrass/, accessed: 2022-10-2.
38. Line notify, https://notify-bot.line.me/zh-TW/, accessed: 2022-10-2.
39. Speed dome, http://www.hertone.com/home/?page_id=161, accessed: 2022-10-2.
40. Hertone company, http://www.hertone.com/home/, accessed: 2022-10-2.
41. Speed dome, http://www.hertone.com/home/?page_id=255, accessed: 2022-10-2.
42. Raspberry Pi, https://en.wikipedia.org/wiki/Raspberry_Pi, accessed: 2022-10-2.
43. Ossrs, https://github.com/ossrs/srs, accessed: 2022-10-2.
44. OpenCV, https://opencv.org/, accessed: 2022-10-2.
45. AWS DynamoDB, https://aws.amazon.com/tw/dynamodb/, accessed: 2022-10-2.
46. WebSocket, https://zh.wikipedia.org/zh-tw/WebSocket, accessed: 2022-10-2.
47. AWS EC2, https://aws.amazon.com/tw/ec2/, accessed: 2022-10-2.
48. Pythonpaho-mqttpackage,https://github.com/eclipse/paho.mqtt.python,accessed:2022- 10-2.
49. 花蓮水產培育所, https://hlab.hl.gov.tw/, accessed: 2022-10-2.