運用3D環境模型之視覺定位方法｜國立中央大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	陳毓琇 Yu-Hsiu Chen
論文名稱：	運用3D環境模型之視覺定位方法 Visual Positioning with 3D Environment Model
指導教授：	黃志煒 Chih-Wei Huang
口試委員:
學位類別：	碩士 Master
系所名稱：	資訊電機學院 - 通訊工程學系 Department of Communication Engineering
論文出版年：	2019
畢業學年度：	107
語文別：	英文
論文頁數：	46
中文關鍵詞：	定位、同時定位與建圖、場景識別、卷積神經網絡
外文關鍵詞：	Localization, SLAM, Place recognition, Convolution Neural Network
相關次數：	點閱：14 下載：0
分享至:	分享至facebook 分享至twitter

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

室內定位至今已發展有一段時間，有很多相關的研究像是場景辨識
以及導航，現有的深度學習定位方法需要大量附有正確相機位置的圖
像，這篇論文主要利用同時定位與建立地圖（SLAM）算法所生成的三
維地圖解決定位問題，我們使用投影方法從3D地圖生成訓練數據，此方
法可以產生在3D地圖中任何地方的圖像，並且帶有準確的位置訊息，我
們也結合了B-CNN[12]所形成的縮放地圖和深度學習解來決定位問題。

Indoor localization has been developed for many years. There are many
related works like scene recognition and navigation. Existing deep learning
positioning methods require a large number of images with the correct camera position. This paper mainly solves the positioning problem by using the
3D map produced from simultaneous localization and mapping (SLAM) algorithm. In our positioning work, we use the projection method to produce
training data from the 3D map. This method can produce any place’s image
in the 3D map included accurate position information. We also combined BCNN [12] to reach a ”zooming map” and deep learning to solve the positioning
problem.

Table of Contents
Introduction 1
1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
3 Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
4 Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Background and Related Work 3
0.1 Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
0.2 Visual Odometry and Add Key Frame . . . . . . . . . . . . . . . . 5
0.3 Loop Closure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
0.4 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
0.5 Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1 Visual Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Branch-Based Classification as Positioning 11
1 Design principle and architecture . . . . . . . . . . . . . . . . . . . . . . . 11
2 Projection Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3 Convolution Neural Network Structure . . . . . . . . . . . . . . . . . . . . 14
3.1 Six Different Cases with VGG16 . . . . . . . . . . . . . . . . . . 15
3.2 Six Different Cases with Branch Convolution Neural Network . . . 16
Implementation and Performance Evaluation 20
1 Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2 Grid Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1 Grids on the 3D Map . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 Grids with B-CNN Structure . . . . . . . . . . . . . . . . . . . . . 22
3 Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4 Training Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.1 Open 3D Map Training Result . . . . . . . . . . . . . . . . . . . . 25
4.2 Engineering Building 3D Map Training Result . . . . . . . . . . . 26
4.3 Validation Accuracy Result . . . . . . . . . . . . . . . . . . . . . . 28
5 Testing Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.1 Presetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.2 Testing Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.3 Positive Samples and Negative Samples of Testing . . . . . . . . . 31
5.4 Branch Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Conclusion and Future Work 34
1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Bibliography 35
                                

Bibliography
[1] Multi session mapping with rtab map tango.
https://github.com/introlab/rtabmap/wiki/Multi-Session-Mapping-with-RTABMap-Tango.
[2] S. S. Ali, A. Hammad, and A. S. Tag Eldien. Mc2ps: Cloud-based 3-d place recognition using map segmentation coordinates points. IEEE Communications Letters,
22(8):1560–1563, Aug 2018.
[3] Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous
driving? the kitti vision benchmark suite. In Conference on Computer Vision and
Pattern Recognition (CVPR), 2012.
[4] L. He, X. Wang, and H. Zhang. M2dp: A novel 3d point cloud descriptor and its
application in loop closure detection. In 2016 IEEE/RSJ International Conference on
Intelligent Robots and Systems (IROS), pages 231–237, Oct 2016.
[5] Mathieu Labbe and Franc¸ois Michaud. Rtab-map as an open-source lidar and visual ´
simultaneous localization and mapping library for large-scale and long-term online
operation: LabbE and michaud. ´ Journal of Field Robotics, 36, 10 2018.
[6] Kin Leong Ho and Paul Newman. Loop closure detection in slam by combining visual
and spatial appearance. Robotics and Autonomous Systems, 54:740–749, 09 2006.
[7] E. Rosten, R. Porter, and T. Drummond. Faster and better: A machine learning approach to corner detection. IEEE Transactions on Pattern Analysis and Machine
Intelligence, 32(1):105–119, Jan 2010.
[8] J. Shotton, B. Glocker, C. Zach, S. Izadi, A. Criminisi, and A. Fitzgibbon. Scene
coordinate regression forests for camera relocalization in rgb-d images. In 2013 IEEE
Conference on Computer Vision and Pattern Recognition, pages 2930–2937, June
2013.
35
[9] Karen Simonyan and Andrew Zisserman. Very deep convolutional networks for largescale image recognition. CoRR, abs/1409.1556, 2015.
[10] F. Walch, C. Hazirbas, L. Leal-Taixe, T. Sattler, S. Hilsenbeck, and D. Cremers. ´
Image-based localization using lstms for structured feature correlation. In 2017 IEEE
International Conference on Computer Vision (ICCV), pages 627–637, Oct 2017.
[11] Zhengyou Zhang. Determining the epipolar geometry and its uncertainty: A review.
International Journal of Computer Vision, 27(2):161–195, 1998.
[12] Xinqi Zhu and Michael Bain. B-CNN: Branch Convolutional Neural Network for
Hierarchical Classification. arXiv e-prints, page arXiv:1709.09890, Sep 2017.

簡易檢索 / 詳目顯示

相關論文