研究生: |
林建利 Jian-Li Lin |
---|---|
論文名稱: |
以點資料處理及深度學習自動化分類堆疊擺放之零件 Automatic Classification of Stacked Parts Using Point Processing and Deep Learning |
指導教授: |
林清安
Ching-An Lin |
口試委員: |
趙振綱
Ching-Kong Chao 陳羽薰 Yu-Hsun Chen |
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: | 2023 |
畢業學年度: | 111 |
語文別: | 中文 |
論文頁數: | 180 |
中文關鍵詞: | 點資料處理 、堆疊零件 、點雲 、深度學習 、機械手臂 、隨機曲放 |
外文關鍵詞: | Point data processing, Deep learning, Random bin picking, Manipulator, Point cloud |
相關次數: | 點閱:193 下載:0 |
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 |
在自動化生產線中,使用機械手臂進行零件分類為主要工作之一,在機械手臂拾取零件前,可透過3D深度學習快速辨識零件種類及數量。然而,當零件堆疊擺放於工作平台上時,難以將零件分割並分別進行辨識,且收集堆疊零件之數據非常耗費時間及人力,為克服以上問題,本論文以點資料處理自動化生成大量堆疊零件之數據,改善數據準備耗時及堆疊零件無法辨識之問題。
本論文透過點資料處理演算法及點雲匹配技術,將零件之點雲以隨機擺放方位進行模擬堆疊,並移除遮蔽點以模擬實際掃描情況,根據此方法生成大量的點雲數據,並自動化將點雲以點為單位新增種類標籤;將生成之數據輸入至PointNet深度學習模型進行訓練,訓練完成之模型平均準確率約為98.1%,而使用該模型辨識真實掃描點雲時,其準確率約為90.24%,兩者之誤差主要出現於零件之間的交界處。
本論文詳述如何使用零件之點雲進行模擬堆疊、3D深度學習之模型及訓練流程,並簡述如何搭配結構光掃描器與機械手臂進行隨機物件夾取,最後對實例驗證之結果進行探討。
In an automated production line, using manipulators to sort parts is one of the main tasks. Before picking up the parts, 3D deep learning can be used to recognize the type and quantity of the parts quickly. However, when the parts are stacked and placed on a work platform, it is difficult to recognize each individual part. The collection of point clouds for all possible combinations of the stacked parts is also very time-consuming and labor-intensive. This thesis approaches the subject by using point data processing to generate a large number of stacked parts automatically, so as to improve the time-consuming of data preparation and unrecognizable problems of stacked parts.
In this thesis, through the point data processing algorithm, the point cloud of the parts is stacked in a random orientation, and the hidden points are removed to simulate the actual scanning situation. According to this method, a large abundant amount of point cloud data is generated and input to PointNet model for deep learning. The average accuracy of the trained model in the test dataset is about 98.1%, and the accuracy in real scanned data is about 90.24%.
In additional to discussing the generation process of stacked point clouds and the training process of 3D deep learning, this thesis also describes the integration of a structured light scanner and a manipulator for gripping and sorting of complex parts, and finally, discusses the results of some case studies.
[1] 劉彥峰(2018),「以機械手臂輔助零件隨機拾取與表面瑕疵檢測之系統開發與應用」,碩士論文,臺灣科技大學機械工程系研究所。
[2] 張仁智(2020),「以機械手臂進行複雜幾何零件之自動化夾取」,碩士論文,國立台灣科技大學機械工程系研究所。
[3] 丁凱庭(2020),「以3D點資料之深度學習搭配機械手臂進行自動化零件分類」,碩士論文,國立台灣科技大學機械工程系研究所。
[4] 賴以衛(2022),「以3D深度學習及點雲匹配技術進行機械手臂自動化複雜零件分類」,碩士論文,國立台灣科技大學機械工程系研究所。
[5] Rusu, R.B., Bradski, G., Thibaux, R. and Hsu, J. (2010), “Fast 3d recognition and pose using the viewpoint feature histogram,” IEEE/RSJ International Conference on Intelligent Robots and Systems, October 18-22, 2010, Taipei, Taiwan, pp. 2155-2162.
[6] Besl, P.J. and McKay, N.D. (1992), “A method for registration of 3-D shapes,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 14, No. 2, pp. 239-256.
[7] Rusu, R.B., Marton, Z.C., Blodow, N. and Beetz, M. (2008), “Persistent point feature histograms for 3D point clouds,” Proceedings of the 10th International Conference on Intelligent Autonomous Systems, July, 2008, Baden, Germany, pp. 119-128.
[8] Rusu, R.B., Blodow, N. and Beetz, M. (2009), “Fast point feature histograms (FPFH) for 3D registration,” IEEE International Conference on Robotics and Automation, May 12-17, Kobe, Japan, 2009, pp. 3212-3217.
[9] Fischler, M.A. and Bolles, R.C. (1981), “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Communications of the ACM, Vol. 24, No. 11, pp. 381-395.
[10] Bogdan, R.R. and Cousins, S. (2011), “3D is here: Point Cloud Library (PCL),” IEEE International Conference on Robotics and Automation, May 9-13, 2011, Shanghai, China, pp. 1-4.
[11] Bentley, J.L. (1975), “Multidimensional binary search trees used for associative searching,” Communications of the ACM, Vol. 18, No. 9, pp. 509-517.
[12] Rusu, R.B., Bradski, G., Thibaux, R. and Hsu, J. (2010), “Fast 3d recognition and pose using the viewpoint feature histogram,” IEEE/RSJ International Conference on Intelligent Robots and Systems, October 18-22, 2010, Taipei, Taiwan, pp. 2155-2162.
[13] Aldoma, A., Vincze, M., Blodow, N., Gossow, D., Gedikli, S. and Rusu, R.B. (2011), “CAD-model recognition and 6DOF pose estimation using 3D cues,” IEEE International Conference on Computer Vision Workshops, November 6-13, 2011, Barcelona, Spain, pp. 585-592.
[14] Sipiran, I. and Bustos, B. (2011), “Harris 3D: A robust extension of the Harris operator for interest point detection on 3D meshes,” The Visual Computer, Vol. 27, No. 11, pp. 963-976.
[15] Liu, M-Y., Tuzel, O., Veeraraghavan, A., Taguchi, Y., Marks, T.K. and Chellappa, R. (2012), “Fast object localization and pose estimation in heavy clutter for robotic bin picking,” International Journal of Robotics Research, Vol. 31, No. 8, pp. 951-973.
[16] 吳政輝(2019),「基於CAD模型之物體姿態辨識及其於機械臂隨機堆疊抓取之應用」,碩士論文,國立交通大學電控工程研究所。
[17] Joseph, R., Divvala, S., Girshick, R. and Farhadi, A. (2016), “You Only Look Once: Unified, real-time object detection,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 27-30, 2016, Las Vegas, NV, USA, pp. 779-788.
[18] Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C-Y. and Berg, A.C. (2016), “SSD: Single Shot MultiBox Detector,” European Conference on Computer Vision, October 11-14, 2016, Amsterdam, The Netherlands, pp. 21-37.
[19] Zeng, A., Yu, K.T., Song, S., Suo, D., Walker, E., Rodriguez, A. and Xiao, J. (2017), “Multi-view self-supervised deep learning for 6D pose estimation in the Amazon Picking Challenge,” IEEE International Conference on Robotics and Automation (ICRA), May 29-June 3, 2017, Singapore, pp. 1383-1386.
[20] Simonyan, K. and Zisserman, A. (2014), “Very deep convolutional networks for large-scale image recognition,” International Conference on Learning Representations, May 7-9, 2014, San Diego, CA, USA.
[21] Wu, Z., Song, S., Khosla, A., Yu, F., Zhang, L., Tang, X. and Xiao, J. (2015), “3D ShapeNets: A deep representation for volumetric shapes,” 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 7-12, 2015, Boston, MV, USA, pp. 1912-1920.
[22] Maturana, D. and Scherer, S. (2015), “VoxNet: A 3d convolutional neural network for real-time object recognition,” IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 28-October 2, 2015, Hamburg, Germany, pp. 922-928.
[23] Qi, C.R., Su, H., Mo, K. and Guibas, L.J. (2017), “PointNet: Deep learning on point sets for 3D classification and segmentation,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), July 21-26, 2017, Honolulu, HI, USA, pp. 652-660.
[24] Rosenblatt, F. (1958), “The perceptron: a probabilistic model for information storage and organization in the brain,” Psychological Review, Vol. 65, No. 6, pp. 386.
[25] Ioffe, S. and Szegedy, C. (2015), “Batch normalization: accelerating deep network training by reducing internal covariate shift,” Proceedings of the 32nd International Conference on International Conference on Machine Learning, July 7-9, 2015, Lille, France, Vol. 37, pp. 448-456.
[26] He, K., Zhang, X., Ren, S. and Sun, J. (2016), “Deep residual learning for image recognition,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 27-30, 2016, Las Vegas, NV, USA, pp. 770-778.
[27] Xia, F., “PointNet.pytorch,” Retrieved from https://github.com/fxi a22/pointnet.pytorch
[28] HP 3D Structured Light Scanner Pro S3, Retrieved from https://physim etrics.com/hp-3d-structured-light-scanner-pro-s3/
[29] EPSON, Retrieved from https://www.epson.eu/en_EU/
[30] SCHUNK,Retrieved from https://schunk.com/ca/en
[31] Point Cloud Library (PCL), Retrieved from https://pointclouds.org/
[32] Open3D, Retrieved from http://www.open3d.org/
[33] PyTorch, Retrieved from https://pytorch.org/