跳到主要內容

臺灣博碩士論文加值系統

(44.220.44.148) 您好!臺灣時間:2024/06/14 10:58
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:林剛永
研究生(外文):LIN,GANG-YONG
論文名稱:基於手勢影像深度學習控制無人自走車
論文名稱(外文):Controlling unmanned self-propelled vehicles based on deep learning of gesture images
指導教授:賴俊吉賴俊吉引用關係
指導教授(外文):LAI,CHUN-CHI
口試委員:賴俊吉林永欽吳文誌
口試委員(外文):LAI,CHUN-CHILIN,YUNG-CHINWU,WEN-CHIH
口試日期:2023-06-24
學位類別:碩士
校院名稱:國立雲林科技大學
系所名稱:電機工程系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2023
畢業學年度:111
語文別:中文
論文頁數:94
中文關鍵詞:手勢辨識圖像轉換深度學習機器人操作系統
外文關鍵詞:gesture recognitionimage conversiondeep learningrobot operating system
相關次數:
  • 被引用被引用:0
  • 點閱點閱:158
  • 評分評分:
  • 下載下載:39
  • 收藏至我的研究室書目清單書目收藏:1
本研究探討利用影像系統辨識人體手部姿態進行判斷,並由手部姿態傳達指令至自走車,電腦影像並利用深度學習取得的成效,使動作辨識能力提高,進而降低以往對於傳感器的廣泛使用,以及增加人機交流。
影像識別利用三原色光(RGB color model)彩色圖像轉換為灰階圖像(Gray scale),再轉為二值化(Binarization),使圖像顯示為黑與白兩種顏色,從而判斷特徵以及收集各式手勢樣本,本研究使用為手勢比出1~5當為控制樣本,利用此五種手勢來控制自走車。收集五種手勢樣本,使用卷機神經網路(Convolutional Neural Network, CNN),將大量數據進行訓練,從而提升影像辨識的準確率。
此研究是將大量數據,進行訓練,提取主要特徵要件,來提升影像辨識度。訓練後的模擬與測試是利用機器人操作系統(Robot Operating System, ROS),從而模擬與控制自走車移動,並利用鏡頭傳入影像,及時判斷手勢變化,並觀察每一刻的辨識狀況。

This study explores the use of image systems to identify human hand postures for judgment, and conveys instructions from hand postures to self-propelled vehicles. Computer imaging and the use of deep learning can improve the ability of motion recognition and reduce the extensive use of sensors in the past. Use, and increase human-computer communication.
Image recognition converts RGB color images into grayscale images, and then converts them into binarization, so that the images are displayed in black and white colors, so as to judge features and collect various gesture samples. This research uses gesture comparison 1~5 are control samples, use these five gestures to control the self-propelled vehicle. Collect five gesture samples, use CNN to train a large amount of data, so as to improve the accuracy of image recognition.
This research is to use a large amount of data for training and extract the main feature elements to improve image recognition. The simulation and test after training is to use the robot operating system operating system to simulate and control the movement of the self-propelled vehicle, and use the camera to transmit images to judge gesture changes in time and observe the recognition status at every moment.

摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
表目錄 vii
圖目錄 viii
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機與目的 2
1.3 研究目標 3
1.4 論文架構 4
第二章 相關文獻探討 5
2.1 影像處理(image processing) 5
2.2 轉換影像為灰階(Transform Image to Gray Level) 5
2.3 二值化(Binarization) 6
2.3.1 OpenCV 中圖像二值方法 8
2.3.2 大津二值化法(OTSU) 9
2.4 卷積神經網路(CNN) 11
2.4.1 卷積層(Convolutional layer) 12
2.4.2 池化(Max Pooling) 14
2.4.3 全連接層(Fully-Connected Layer) 15
第三章 軟體與程式介紹 16
3.1 VMware Workstation Player 16
3.2 Ubuntu 17
3.3 ROS(Robot Operating System) 18
3.3.1 Gazebo 18
3.3.2 RViz 19
3.3.3 ROS節點通訊方式 20
3.4 YOLO 23
3.4.1 物體偵測 24
3.4.2 YOLO訓練 25
3.4.3 版本差異 26
3.4.4 YOLOv5 27
第四章 研究方法 31
4.1 地圖建立 31
4.1.1 TurtleBot3房子地圖 32
4.2 VMware Workstation Player相機設置 33
4.3 二值化影像辨識 34
4.3.1 手勢數據收集 34
4.3.2 手勢訓練 36
4.3.3 手勢輸出測試 40
4.4 YOLO影像辨識 42
4.4.1 數據收集 42
4.4.2 手勢訓練 42
4.4.3 手勢輸出測試 46
4.5 機器人座標設定 48
4.6 ROS上手勢控制機器人 51
第五章 研究成果 57
5.1 二值化影像辨識 57
5.1.1 影像顯示 58
5.1.2 收集樣本 59
5.1.3 結果呈現 64
5.2 YOLO影像辨識 67
5.2.1 影像辨識提升 67
5.2.2 結果呈現 69
5.3 實體呈現 71
第六章 結論與未來展望 75
6.1 結論 75
6.1.1 二直化影像辨識 75
6.1.2 YOLO架構辨識 76
6.2 未來研究與展望 76
參考文獻 78


[1]García, D. H., Esteban, P. G., Lee, H. R., Romeo, M., Senft, E., & Billing, E. (2019, March). Social robots in therapy and care. In 2019 14th ACM/IEEE International Conference on Human-Robot Interaction (HRI) (pp. 669-670). IEEE.
[2]LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278-2324.
[3]Zheng, L., Ai, P., & Wu, Y. (2020, September). Building recognition of UAV remote sensing images by deep learning. In IGARSS 2020-2020 IEEE International Geoscience and Remote Sensing Symposium (pp. 1185-1188). IEEE.
[4]Song, S., Yan, D., & Xie, Y. (2018, March). Design of control system based on hand gesture recognition. In 2018 IEEE 15th International Conference on Networking, Sensing and Control (ICNSC) (pp. 1-4). IEEE.
[5]Liu, J., Furusawa, K., Tateyama, T., Iwamoto, Y., & Chen, Y. W. (2019, September). An improved hand gesture recognition with two-stage convolution neural networks using a hand color image and its pseudo-depth image. In 2019 IEEE international conference on image processing (ICIP) (pp. 375-379). IEEE.
[6]ROS 資料參考網址: https://zh.wikipedia.org/zh-tw/%E6%A9%9F%E5%99%A8%E4%BA%BA%E4%BD%9C%E6%A5%AD%E7%B3%BB%E7%B5%B1
[7]Yin He, & Song Jiasheng. (2019). Control and monitoring method of intelligent car based on ROS system. Information Recording Materials , 9 .
[8]數位像素資料參考網址: https://zh.wikipedia.org/zh-tw/%E6%95%B0%E5%AD%97%E5%9B%BE%E5%83%8F
[9]Saravanan, C. (2010, March). Color image to grayscale image conversion. In 2010 Second International Conference on Computer Engineering and Applications (Vol. 2, pp. 196-199). IEEE.
[10]二值化過程參考http://140.134.32.129/scteach/mech/text/yin03_12.html
[11]閾值Thresholding參考https://hackmd.io/@Youwe/SkBHZmcrI
[12]Dos Anjos, A., & Shahbazkia, H. R. (2008). Bi-level image thresholding. Biosignals, 2, 70-76.
[13]大津演算法參考 https://zh.wikipedia.org/zh-tw/%E5%A4%A7%E6%B4%A5%E7%AE%97%E6%B3%95
[14]OTSU算法參考 https://blog.csdn.net/bosszhao20190517/article/details/105837566
[15]卷積神經網路參考 https://zh.wikipedia.org/zh-tw/%E5%8D%B7%E7%A7%AF%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C
[16]Albawi, S., Mohammed, T. A., & Al-Zawi, S. (2017, August). Understanding of a convolutional neural network. In 2017 international conference on engineering and technology (ICET) (pp. 1-6). Ieee.
[17]Convolutional Neural Network, CNN 參考 http://elmer-storage.blogspot.com/2018/07/cnn-convolutional-neural-network-cnn.html
[18]O'Shea, K., & Nash, R. (2015). An introduction to convolutional neural networks. arXiv preprint arXiv:1511.08458.
[19]黃婉蓉, 何凱, 劉坤, & 高聖楠. (2020). 基於注意力機制的手寫體中文字符識別.激光與光電子學進展, 57 (8), 081002.
[20]VMware Workstation Player參考資料: https://zh.wikipedia.org/zh-tw/VMware_Workstation_Player
[21]Ubuntu參考資料: https://zh.wikipedia.org/zh-tw/Ubuntu
[22]ROS參考資料: https://zh.wikipedia.org/zh-tw/%E6%A9%9F%E5%99%A8%E4%BA%BA%E4%BD%9C%E6%A5%AD%E7%B3%BB%E7%B5%B1
[23]Gazebo&RViz參考資料: https://ithelp.ithome.com.tw/articles/10248650
[24]RViz 數據類型參考: https://www.guyuehome.com/36509
[25]ROS Node參考資料: https://hackmd.io/@ncrl-10/ByOwDYF6t
[26]ROS Topic相關指令參考: https://ithelp.ithome.com.tw/articles/10268791
[27]Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 779-788).
[28]深度學習-物測參考: https://chih-sheng-huang821.medium.com/%E6%B7%B1%E5%BA%A6%E5%AD%B8%E7%BF%92-%E7%89%A9%E4%BB%B6%E5%81%B5%E6%B8%AC-you-only-look-once-yolo-4fb9cf49453c
[29]Anchor box參考: https://zhuanlan.zhihu.com/p/63024247
[30]Liu, C., Tao, Y., Liang, J., Li, K., & Chen, Y. (2018, December). Object detection based on YOLO network. In 2018 IEEE 4th information technology and mechatronics engineering conference (ITOEC) (pp. 799-803). IEEE.
[31]Shinde, S., Kothari, A., & Gupta, V. (2018). YOLO based human action recognition and localization. Procedia computer science, 133, 831-838.
[32]YOLO v1~v5版本差異參考: https://u9534056.medium.com/%E7%9B%AE%E6%A8%99%E6%AA%A2%E6%B8%AC-yolo-v1-v5-%E5%85%A8%E7%89%88%E6%9C%AC%E5%B7%AE%E7%95%B0-c648cc5e49f1
[33]Redmon, J., & Farhadi, A. (2017). YOLO9000: better, faster, stronger. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 7263-7271).
[34]Redmon, J., & Farhadi, A. (2018). Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767.
[35]Bochkovskiy, A., Wang, C. Y., & Liao, H. Y. M. (2020). Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934.
[36]YOLOv5參考資料: https://roboflow.com/model/yolov5
[37]YOLOv5參考資料: https://pytorch.org/hub/ultralytics_yolov5/
[38]YOLOv5架構資料: https://github.com/ultralytics/yolov5/releases
[39]MediaPipe資料參考:https://ithelp.ithome.com.tw/articles/10265041
[40]XGBoost資料參考: https://medium.com/@techynilesh/xgboost-algorithm-explained-in-less-than-5-minutes-b561dcc1ccee

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊