(3.215.180.226) 您好!臺灣時間:2021/03/09 03:34
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:::

我願授權國圖
: 
twitterline
研究生:謝承哲
研究生(外文):XIE, CHENG-ZHE
論文名稱:以聯合相關濾波器與軌跡預測 之即時目標物追蹤
論文名稱(外文):Real-Time Object Tracking with Joint Correlation Filters and Trajectory Prediction
指導教授:黃正民黃正民引用關係
指導教授(外文):HUANG, CHENG-MING
口試委員:連豐力簡忠漢練光祐江明理黃正民
口試委員(外文):LIAN, FENG-LICHIEN, CHUNG-HANLIEN, KUANG-YUCHIANG, MING-LIHUANG, CHENG-MING
口試日期:2020-07-28
學位類別:碩士
校院名稱:國立臺北科技大學
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:2020
畢業學年度:108
語文別:中文
論文頁數:88
中文關鍵詞:視覺追蹤相關濾波器卡爾曼濾波器模型更新機制旋轉估測軌跡預測
外文關鍵詞:Visual TrackingCorrelation FiltersKalman FilterUpdate StrategyRotation EstimationTrajectory Prediction
相關次數:
  • 被引用被引用:0
  • 點閱點閱:36
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
影像處理技術之相關應用在這幾年蓬勃發展,而視覺追蹤就是其中一項重要的研究議題。本論文將提出一個具即時性之穩健視覺追蹤系統,能解決在平面旋轉、快速位移等場景估測不穩定之問題。本系統結合位置、旋轉、尺度等三種相關濾波器之狀態估測,來建構聯合相關濾波器,成為新的追蹤架構。
由於相關濾波器在位置估測上受到搜尋範圍限制,因此本篇論文將實驗較適合系統之搜尋範圍大小,並結合卡爾曼濾波器預測目標物之路徑,增加多個採樣點,解決搜尋範圍太小之限制,防止目標物追蹤在快速位移下失誤的問題。而對於旋轉估測在非平面旋轉場景會產生估測錯誤之問題,本篇論文採用長短期記憶網路來預測目標物之旋轉角度,限制非預期之旋轉角度變化。尺度估測則因為受到樣本尺度影響,在效率上特性不一樣,所以在金字塔架構提取樣本與進行旋轉等兩個先後處理順序間進行動態切換,以達到較高效率的採樣策略。
在視覺追蹤初始化後,適當的更新目標物模型策略能夠適應更多種不同的環境,同時又能避免目標物模型在追蹤失誤時受到汙染。在此,相關濾波器的響應圖可視為追蹤置信度依據,本研究將利用峰值評估指標APCE和PSR為支撐向量機分類使用之特徵,決定是否更新模型。實驗結果將呈現本研究採用之方法效果,並與多種追蹤方法比較準確度與速度。最後,本論文將在OTB100數據集上進行實驗,以呈現本文追蹤系統之準確性和即時性。

The applications of image processing have been booming in these years, and visual object tracking is one of the most challenging problems in computer vision. In this paper, we propose a real-time robust visual tracking system which can solve the problem of unstable estimation in scenes such as plane rotation and rapid displacement. The proposed new tracking architecture is a joint correlation filter which combines three correlation filters such as position, rotation and scale to estimate the state of target.
Since the correlation filter is restricted by the size of search region in position estimation, we will experiment to select more suitable size for the search range of the system. Combined with the Kalman filter to predict the path of the target object, multiple sampling points are added to solve the limitation of the search range is too small and prevent the target tracking error under fast displacement. Regarding the problem that rotation estimation will produce estimation errors in non-planar rotation scenes, we utilize the long short-term memory network to predict the rotation angle of the target and limit unexpected rotation angle changes. The scale estimation is affected by the sample scale and has different characteristics in processing efficiency. Therefore, dynamic switching is performed between the two sequential processing sequences of extracting samples and performing rotation in the pyramid structure to achieve a sampling strategy with higher efficiency.
In addition, a proper updated strategy after initialization is important for tracking arbitrary object and avoiding model degeneration. For the issue, the response map of correlation filter during tracking can be considered as the tracking confidence. We also exploit the confidence metrics, including average peak-to correlation energy (APCE) and peak to sidelobe ratio (PSR), as features for support vector machine to be update strategy. The experimental results on the OTB100 dataset show the precision and real-time performance of the proposed algorithms in this paper.
摘要 i
ABSTRACT iii
致謝 v
目錄 vi
表目錄 ix
圖目錄 x
1. 第一章 緒論 1
1.1. 前言 1
1.2. 研究動機 3
1.3. 研究成果與貢獻 6
1.4. 系統流程 7
1.5. 論文架構 8
2. 第二章 相關文獻 9
2.1. 視覺追蹤 9
2.1.1. 基於特徵匹配追蹤 10
2.1.2. 基於粒子濾波器追蹤 11
2.1.3. 基於卷積神經網路追蹤 11
2.1.4. 其他機器學習—線上結構化SVM 13
2.1.5. 相關濾波器追蹤 14
2.2. 相關濾波器模型更新機制 16
2.3. 平面旋轉偵測 18
3. 第三章 聯合相關濾波器 19
3.1. 目標物狀態描述及流程 20
3.2. 相關濾波器 21
3.2.1. 位置估測 21
3.2.2. 旋轉估測 26
3.2.3. 尺度估測 29
3.2.4. 聯合狀態估測 31
3.3. 攝影機運動估測 32
4. 第四章 更新機制與狀態預測 34
4.1. 更新機制 34
4.1.1. 相關濾波器模型更新 34
4.1.2. 支持向量機 36
4.1.3. 追蹤置信度指標建模 38
4.2. 位置預測 41
4.2.1. 卡爾曼濾波器 41
4.2.2. 運動模型 42
4.3. 旋轉預測 45
4.3.1. 長短期記憶神經網路 45
4.3.2. 預測建模與限制角度機制 47
4.3.3. 旋轉角度重置機制 49
5. 第五章 實驗結果 50
5.3. 搜尋範圍比較 52
5.4. 加入卡爾曼路徑預測比較 52
5.5. 加入旋轉量比較 54
5.5.1. 位置估測比較 54
5.5.2. 尺度估測比較 55
5.5.3. 旋轉樣本效率實驗 56
5.6. 加入旋轉預測限制旋轉變化 58
5.7. 更新機制比較 59
5.8. 遮蔽場景下之旋轉估測效果分析 62
5.9. 光流法補償效果分析 64
5.10. 運算時間比較 65
5.11. 多種追蹤方法結果比較 67
5.11.1. 實驗一:快速位移與複雜背景之視覺追蹤 70
5.11.2. 實驗二:快速位移與形變之視覺追蹤 72
5.11.3. 實驗三:快速位移與運動模糊之視覺追蹤 74
5.11.4. 實驗四:平面與非平面旋轉之視覺追蹤 76
5.11.5. 實驗五:畫面旋轉與尺度變化之視覺追蹤 78
5.11.6. 實驗六:長期遮蔽之視覺追蹤 80
5.12. 完整數據結果分析 83
6. 第六章 結論與未來展望 84
6.1. 結論 84
6.2. 未來展望 85
參考文獻 86

[1] Wu, Yi, Jongwoo Lim, and Minghsuan Yang., “Online Object Tracking: A Benchmark,” in 2013 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2013.
[2] 比高, “ 8-個遙控小秘技助你解決空拍機飛行問題” [Online]. Available:
https://dronesplayer.com/aerial-photography
[3]“DJI™ 官網 | DJI.com‎” [Online]. Available:
https://store.dji.com/zh-tw/shop/phantom-series?from=menu_icon
[4] cinopanacea, “ Phantom 4, DJI GO app on iPhone 6s Plus: Active Track.” [Online]. Available:https://www.youtube.com/watch?v=56XOp1YqMR8
[5] J. F. Henriques, R. Caseiro, P. Martins, and J. Batista, “High-speed tracking with kernelized correlation filters,” in IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015.
[6] Nebehay, G., and Pflugfelder., R., “Consensus-based matching and tracking of keypoints,” In Winter Conference on Applications of Computer Visio, 2014.
[7] Ross, D. A.; Lim, J.; Lin, R.-S.; and Yang, M.-H., “Incremental learning for robust visual tracking,” in IJCV, 2008.
[8] Ji, H., “Real time robust l1 tracker using accelerated proximal gradient approach,” in 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012.
[9] M. Talha and R. Stolkin, “Particle Filter Tracking of Camouflaged Targets by Adaptive Fusion of Thermal and Visible Spectra Camera Data,” in IEEE Sensors Journal, vol. 14, no. 1, pp. 159–166, Jan. 2014.
[10] Nam, H., and Han, B., “Learning multi-domain convolutional neural networks for visual tracking,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
[11] Bertinetto, L.; Valmadre, J.; Henriques, J.; Vedaldi, A.; and Torr, P. H., “Fullyconvolutional siamese networks for object tracking,” in 2016 ECCV Workshop.
[12] D. S. Bolme, J. R. Beveridge, B. Draper, Y. M. Lui et al., “Visual object tracking using adaptive correlation filters,” in 2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010.
[13] Tianzhu Zhang, Changsheng Xu, Ming-Hsuan Yang., “Multi-Task Correlation Particle Filter for Robust Object Tracking,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
[14] Yang Li, Jianke Zhu, Steven C.H. Hoi, Wenjie Song, Zhefeng Wang, Hantang Liu., “Robust Estimation of Similarity Transformation for Visual Object Tracking,” in 2019 The National Conference on Artificial Intelligence (AAAI), 2019.
[15] M. Danelljan, F. Shahbaz Khan, M. Felsberg, and J. van de Weijer, “Adaptive color attributes for real-time visual tracking,” in 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2014.
[16] Ming Tang, Bin Yu, Fan Zhang, Jinqiao Wang, “High-speed Tracking with Multi-kernel Correlation Filters,” in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
[17] Martin Danelljan, Goutam Bhat, Fahad Shahbaz Khan, Michael Felsberg, “ECO: Efficient Convolution Operators for Tracking,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
[18] S. Hare, A. Saffari, and P. H. Torr, “Struck: Structured output tracking with kernels,” in 2011 IEEE International Conference on Computer Vision(ICCV), 2011.
[19] Chao Ma, Xiaokang Yang, Chongyang Zhang, and Ming-Hsuan Yang, “Long-term Correlation Tracking,” in 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015.
[20] J. Henriques, R. Caseiro, P. Martins, and J. Batista., “Exploiting the Circulant Structure of Tracking-by-detection with Kernels,” in Proceedings of the European Conference on Computer Vision (ECCV), 2012.
[21] Martin Danelljan, Andreas Robinson, Fahad Khan, Michael Felsberg., “Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking,” in Proceedings of the European Conference on Computer Vision (ECCV), 2016.
[22] Martin Danelljan, Gustav Häger, Fahad Shahbaz Khan, Michael Felsberg, “Discriminative Scale Space Tracking,” p.p.1561-1575, in 2017 IEEE Transactions on Pattern Analysis and Machine Intelligence.
[23] M. Tang and J. Feng., “Multi-kernel correlation filter for visual tracking,” in 2015 IEEE International Conference on Computer Vision (ICCV), 2015.
[24] Wang M, Liu Y, Huang Z., “Large Margin Object Tracking with Circulant Feature Maps,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
[25] Li, Y., and Zhu, J., “A scale adaptive kernel correlation filter tracker with feature integration,” in ECCV Workshops, 2014.
[26] B. Sch¨olkopf and A. Smola., “Learning with Kernels,” MIT press Cambridge, MA., 2002.
[27] Chih-Chung Chang, Chih-Jen Lin, “LIBSVM -- A Library for Support Vector Machines,” [Online]. Available: https://www.csie.ntu.edu.tw/~cjlin/libsvm/
[28] Oinkina, “Understanding LSTM Networks,” [Online]. Available:
http://colah.github.io/posts/2015-08-Understanding-LSTMs/
[29] A. Sherstinsky, “Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) Network,” arXiv:1808.03314 [cs], Aug. 2018.
[30] Andrej Karpathy, “The Unreasonable Effectiveness of Recurrent Neural Networks.” [Online]. Available: http://karpathy.github.io/2015/05/21/rnn-effectiveness
[31] Boris Babenko, Ming-Hsuan Yang, and Serge Belongie, “Visual tracking with online multiple instance learning,” in 2009 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2009, pp. 983–990.
[32] Z. Kalal, J. Matas, and K. Mikolajczyk., “P-N Learning: Bootstrapping Binary Classifiers by Structural Constraints,” in 2010 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010.
[33] K. Zhang, L. Zhang, and M.-H. Yang, “Real-time compressive tracking,” in European Conference on Computer Vision (ECCV), 2012, pp. 864–877.
[34] Yulia Makmur, “第六章,數字信號處理技術工程測試技術基礎本章學習要求,” [Online]. Available: https://slidesplayer.com/slide/15974003/
[35] opencv dev team, “Geometric Image Transformations,” [Online]. Available: https://docs.opencv.org/2.4/modules/imgproc/doc/geometric_transformations.html
[36] Bo-Wei Jianga, Cheng-Ming Huang, “Robust Real-Time Visual Tracking by Adaptively Fusing Color Camera and Thermal Camera,” in 2017 IEEE National Symposium on Systems Science and Engineering (NSSSE), 2017.

電子全文 電子全文(網際網路公開日期:20250825)
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔