本研究針對蘭花芽株培養，開發能自動以機器手臂夾取蘭花芽株出瓶和自動將蘭花芽株莖切割後的小芽株放入於瓶內的智慧自動化技術，結合2D和3D的視覺技術，並結合六軸串聯式機械手臂進行物件夾取。上述技術結合相機基礎理論、雙目視覺理論、Perspective-n-Point (PnP)來進行相機定位、機器人學和深度學習來訓練2D物件辨識。
針對自動夾取蘭花芽株出瓶系統，因在培養中的瓶子無法使深度相機有效建模，本研究使用雙目相機來克服此問題以達成從瓶中夾取蘭花工作。針對將蘭花芽株放入培養瓶系統，因不會有杯壁干擾3D建模，故將深度相機固定於機械手臂上進行影像建模。
為確保判斷芽株的頭部及根部位置正確，使用深度學習判斷方位，再進行夾取動作，夾取的過程中，受到夾具設計以及機器人運動學的影響，需分成兩部分進行兩段式放入，第一段過程為辨識物件的頭尾，再夾取並依序將其放入固定架上，第二段過程為規劃機械手臂放入路徑，將固定架上的蘭花芽株依序放入瓶中。

This research aims to develop robot arm-based autonomous orchid bud operation techniques including picks orchid buds out of the bottle and place the orchid buds after fixing operation back to the bottle. The techniques implement, both 2D and 3D vision systems and the use of a six-axis robot arm used for object grasping. The techniques used include basic camera theory, binocular vision theory, Perspective-n-Point (PnP) for camera positioning, robotics and2D object recognition using deep learning.
For the automatic grasping and picking of orchid buds out of the bottle, because the general depth camera modeling method often based on laser-assisted positioning and modeling, the depth camera cannot effectively model the transparent bottle used to contain orchid buds. This research uses a binocular camera to overcome this problem and achieve the task of grasping the orchid buds from the bottle. In the placing orchid buds back bottle system, the transparent bottle wall will not interfere with the 3D modeling. Therefore, the depth camera is used to attached on the robot arm for image modeling.
To ensure the correct positioning of the head and roots of the orchid buds, the deep learning algorithm is used to determine the correct position, and then provide information for the grasping action. The grasping process is divided into two parts. The first stage is to identify the head and tail of the orchid buds, then grab and place it in the fixed rack in sequence. The second stage of the process is to pick the items and place back the bottle in order.

1.Forsyth, D.A. and J. Ponce, A modern approach. Computer vision: a modern approach, 2003: p. 88-101.
2.Heikkila, J. and O. Silven. A four-step camera calibration procedure with implicit image correction. in Computer Vision and Pattern Recognition, 1997. Proceedings., 1997 IEEE Computer Society Conference on. 1997. IEEE.
3.Zhang, Z., A flexible new technique for camera calibration. IEEE Transactions on pattern analysis and machine intelligence, 2000. 22.
4.針孔投影圖，取自<https://blog.techbridge.cc/2018/04/22/intro-to-pinhole-camera-model/>
5.針孔成像原理，取自<https://abcd40404.github.io/2018/08/08/camera-frame-world-frame/>
6.Spong, M.W., S. Hutchinson, and M. Vidyasagar, Robot modeling and control. Vol. 3. 2006: Wiley New York.
7.Duane, C.B., Close-range camera calibration. Photogramm. Eng, 1971. 37(8): p. 855-866.
8.Banglei Guan, Yang Shang, and Qifeng Yu, "Planar self-calibration for stereo cameras with radial distortion," Appl. Opt. 56, 9257-9267 (2017)
9.相機形變參數介紹，取自<https://silverwind1982.pixnet.net/blog/post/153218861>
10.Fryer, J.G. and D.C. Brown, Lens distortion for close-range photogrammetry. Photogrammetric engineering and remote sensing, 1986. 52(1): p. 51-58.
11.三角測量介紹，取自<https://zh.wikipedia.org/wiki/%E4%B8%89%E8%A7%92%E6%B8%AC%E9%87%8F >
12.Gao, Xiao-Shan, et al. "Complete solution classification for the perspective-three-point problem." IEEE transactions on pattern analysis and machine intelligence 25.8 (2003): 930-943.
13.Horaud, R., et al. An analytic solution for the perspective 4-point problem. in Computer Vision and Pattern Recognition, 1989. Proceedings CVPR'89., IEEE Computer Society Conference on. 1989. IEEE.
14.Zhi, Lihong, and Jianliang Tang. "A complete linear 4-point algorithm for camera pose determination." AMSS, Academia Sinica 21 (2002): 239-249.
15.Quan, Long, and ZhongdanLan. "Linear n-point camera pose determination." IEEE Transactions on pattern analysis and machine intelligence 21.8 (1999): 774-780.
16.Harris, C. and M. Stephens. A combined corner and edge detector. in Alvey vision conference. 1988. Citeseer.
17.Steven J. Miller. The Method of Least Squares, Department of Mathematics and Statistics, Williams College, Williamstown, Massachusetts
18.Tsai, R.Y. and R.K. Lenz. Real time versatile robotics hand/eye calibration using 3D machine vision. in Robotics and Automation, 1988. Proceedings., 1988 IEEE International Conference on. 1988. IEEE.
19.Dornaika, F. and R. Horaud, Simultaneous robot-world and hand-eye calibration. IEEE transactions on Robotics and Automation, 1998. 14(4): p. 617-622.
20.Bayro-Corrochano, E., K. Daniilidis, and G. Sommer, Motor algebra for 3D kinematics: The case of the hand-eye calibration. Journal of Mathematical Imaging and Vision, 2000. 13(2): p. 79-100
21.Strobl, K.H. and G. Hirzinger. Optimal hand-eye calibration. in Intelligent Robots and Systems, 2006 IEEE/RSJ International Conference on. 2006. IEEE.
22.機器人學圖，取自<https://zh-tw.coursera.org/lecture/robotics1/3-3-dhbiao-da-fa-3LjyX>
23.雙目視覺，取自<https://blog.csdn.net/qq_16181387/article/details/78442758?utm_medium=distribute.pc_relevant_download.none-task-blog-searchfrombaidu-4.nonecase&depth_1-utm_source=distribute.pc_relevant_download.none-task-blog-searchfrombaidu-4.nonecas>
24.yolov4深度學習，取自<https://medium.com/ching-i/yolo%E6%BC%94%E9%80%B2-3-yolov4%E8%A9%B3%E7%B4%B0%E4%BB%8B%E7%B4%B9-5ab2490754ef>