研究生: |
柯霽祐 JI-YOU KE |
---|---|
論文名稱: |
以Android行動裝置及雲端影像處理為基礎的主從式機器人 A Client-Server Robot Based on Android Device and Cloud Image Processing |
指導教授: |
高維文
Wei-Wen Kao |
口試委員: |
黃緒哲
Shiuh-Jer Huang 林紀穎 Chi-Ying Lin |
學位類別: |
碩士 Master |
系所名稱: |
工程學院 - 機械工程系 Department of Mechanical Engineering |
論文出版年: | 2016 |
畢業學年度: | 104 |
語文別: | 中文 |
論文頁數: | 52 |
中文關鍵詞: | 影像處理 、TLD 、物體追蹤 、Android 、物聯網 、雙輪運動機器人 |
外文關鍵詞: | image processing, TLD, object tracking, Android, IoT, 2dw robot. |
相關次數: | 點閱:241 下載:7 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
智慧型手機人手一機,將人們帶入了物聯網時代。跟隨著時代的進步,演算的成本不斷下降,高品質的攝影機也越賣越便宜,而有許多高度依賴視覺的機器人不斷的上市。但對於機器人內建的處理器而言,影像處理的工作還是過於繁重,因此將繁重的運算交由雲端科技解決,是更理想的解決方式。
本論文題出自行建置一主從式系統架構,其中包含一台雙輪運動機器人,以及一個雲端的影像處理器。雙輪運動機器人上搭載了一台 Android 手機,於手機上運行NIO 伺服器,以傳送串流影像至雲端,以及接收雲端運算的結果,並以結果決定機器人的行動。雲端的影像處理器採用 TLD 演算架構,能夠追蹤任意物體,同時線上學習其物體辨識模型。雲端處理器也是於 Android 系統上執行。
The idea of the internet of Things (IoT) has been a heated topic for scholars and engineers since smartphone gained popularity among almost every person at every age. As the cost of both computing and high-quality camera are reducing, more and more robots with higher dependency on visual sensor have come to the market.Yet the loading of image processing for a built-in computer is still too heavy. Thus, a solution of combining cloud computing technology would be more ideal.
This research is based on a client-server architecture system we built, consisting of a 2WD (two-wheel drive) robot and a cloud image processor. The system is then modified to set up an IoT environment, serving as a backbone for the robot’s tracking function. The mobile robot in the system is equipped with an Android phone which is used as a server to transfer image stream to the cloud, and receiving the processed result, then determine the motion of robot. The image processor on the cloud is based on the TLD (Tracking-Learning-Detection) framework, it is able to track an arbitrary object and built the model for detection online. The image processor also run on Android OS.
1. First Android Smartphone Satellite Now Orbiting Earth. 2015; Available from: http://www.space.com/19996-first-android-smartphone-satellite-now-orbiting-earth-video.html.
2. Goransson, A., Android 高效能多執行緒. 2015/05/26: O'Reilly
3. 朱建亮, 網路型機器人的設計與單相機 LSD 同步定位與建圖的實現, in 機械工程系. 2015, 國立臺灣科技大學: 台北.
4. 達明翰, 基於HOG與TLD方法實現行人的偵測與追蹤, in 機械工程系. 2015, 國立臺灣科技大學: 台北.
5. 鄭禮逸, 基於鼻部特徵之頭部姿態即時追蹤機器人之實現, in 機械工程系. 2015, 國立臺灣科技大學: 台北.
6. G. Allen, R.Y.D.X., and J. S. Jin. , Object Tracking Using CamShift Algorithm and Multiple Quantized Feature Spaces. 2006, University of Sydney.
7. Kanade, B.D.L.a.T. An iterative image registration technique with an application to stereo vision. in Proceeding of the 7th International Joint Conference on Artificial Intelligence. 1981. British Clumbia.
8. Schunk., B.K.P.H.a.B., Determining Optical Flow., in Artificial Intelligence. 1981.
9. Williams, O., A. Blake, and R. Cipolla, Sparse Bayesian learning for efficient visual tracking. Ieee Transactions on Pattern Analysis and Machine Intelligence, 2005. 27(8): p. 1292-1304.
10. Kalal, Z., K. Mikolajczyk, and J. Matas, Tracking-Learning-Detection. Ieee Transactions on Pattern Analysis and Machine Intelligence, 2012. 34(7): p. 1409-1422.
11. USB Host. USB Host and Accessory [cited 2016; Available from: http://developer.android.com/guide/topics/connectivity/usb/host.html.
12. Hitchens, R., Java Nio. O'REILLY.
13. Transmitting Network Data Using Volley. 2016; Available from: http://developer.android.com/training/volley/index.html.
14. Wang, T.P.B.a.A.-I. Linux Kernel & Device Driver Programming. 2010; Available from: http://www.cs.fsu.edu/~baker/devices/.
15. Zdenek Kalal, K.M., Jiri Matas, Forward-Backward Error: Automatic Detection of Tracking Failures, in 2010 International Conference on Pattern Recognition. 2010.
16. 朱慶峰, 以適應性 KLT 為特徵之移動物定位系統, in 多媒體工程研究所. 2011, 國立交通大學.
17. Birchfield, S., KLT: An Implementation of the Kanade-Lucas-Tomasi Feature Tracker. 1997.
18. Kanade, B.D.L.a.T., An Iterative Image Registration Technique with an Application to Stereo Vision, in International Joint Conference on Artificial Intelligence. 1981. p. 674-679.
19. Shi, J.B. and C. Tomasi, Good Features to Track. 1994 Ieee Computer Society Conference on Computer Vision and Pattern Recognition, Proceedings, 1994: p. 593-600.
20. Kalal, Z., J. Matas, and K. Mikolajczyk, P-N Learning: Bootstrapping Binary Classifiers by Structural Constraints. 2010 Ieee Conference on Computer Vision and Pattern Recognition (Cvpr), 2010: p. 49-56.