臺灣博碩士論文加值系統

本論文提出一個透過彩色攝影機結合熱像儀的影像追蹤視覺系統，能夠有效地對目標物進行追蹤。由於在追蹤的過程中，有許多外在的因素影響，可能導致無法準確成功的對目標定位與追蹤，如：攝影機本身劇烈的晃動下，會導致影像畫面中的目標可能會模糊以及失焦;目標物移動過程中，隨著視角的不同造成其外型或外貌改變；以及當目標物被一些外在環境中的遮蔽物阻擋瞄準視線，無法確認前方物體是需鎖定的目標。綜合以上所述，本文將建立一個視覺追蹤系統，透過一群具有權重的粒子，描述欲追蹤目標物之運動狀態，以有效地對目標物狀態進行估測。
由於所追蹤目標為行人或車輛，利用目標本身溫度顯示於熱紅外線攝影機，對視覺追蹤將融合彩色可見光攝影機與熱紅外線攝影機感測之影像，藉由適應性多種重要性採樣之粒子濾波器產生目標物可能所在之位置。並在於彩色可見光攝影機提出多種特徵經由粒子濾波器進行目標物相似度比較，提升粒子濾波器估測目標物的準確性。然而當目標物移動過程中，被外在環境有所遮蔽時，加入稀疏表示法計算粒子權重分數提高粒子的相似性以克服目標物外貌改變與被遮蔽等難題，成功穩定對目標物追蹤。除此之外，也將攝影機固定於一雙軸旋轉馬達平台上，利用即時視覺追蹤目標物的結果，控制雙軸旋轉馬達平台以將目標物持續鎖定於攝影機影像畫面中央。

In this paper, a visual tracking system by using the particle filter with a monocular color camera and a thermal camera is proposed to track the target. The thermal camera, which can observe the heat originated from the target such as the human body or vehicle, collaborates with the color camera to track the target in the cluttered environment or under occlusion. First, the extrinsic parameters between the color camera and the thermal camera are calibrated to estimate the homography matrix between two cameras’ coordinates. An adaptive multiple importance sampling scheme is then developed to efficiently generate the particle hypotheses by dynamically fusing the target clues in the color camera and the thermal camera. These hypotheses are verified by evaluating the target similarity on the distributions of color and significant edge in the color image. Moreover, the sparse appearance model is combined into the likelihood evaluation to improve the tracking precision when the target might be occluded. In addition, the camera will also be fixed on a dual-axis rotary motor platform, with visual target tracking results, the biaxial rotary motor control platform to continue to track the target in the center of the camera image screen. Finally, the proposed approaches have been validated in several scenes to present the tracking performance.

中文摘要 i
英文摘要 ii
誌謝 iv
目錄 vi
表目錄 viii
圖目錄 ix
第一章 緒論 1
1.1 前言 1
1.2 研究動機 1
1.3 相關文獻 2
1.4 研究成果與貢獻 3
1.5 硬體設備 4
1.6 論文架構 6
第二章 融合彩色攝影機與熱像儀目標物影像追蹤 7
2.1目標物之狀態描述 7
2.2熱像儀影像與彩色攝影機外部參數校正 8
2.3適應性多個重要性採樣 9
2.3.1熱像儀目標物區域偵測 11
第三章 目標物影像特徵比對 15
3.1目標物之參考模板建立 16
3.2 粒子相似度計算 17
3.2.1目標物追蹤之模板相似度計算 17
3.2.2追蹤目標物之稀疏表示 18
第四章 地面平台與目標物影像對準控制 21
4.1旋轉馬達控制 21
4.2預測旋轉馬達角度 24
第五章 實驗結果 25
5.1室內行人 25
5.2室內遮蔽 27
5.3戶外行人 29
5.4戶外車子 31
5.5夜晚 34
5.6稀疏表示與為未稀疏表示 40
5.7戶外遠方行人 41
5.8戶外遠方車輛 42
5.9本文方法與其他學者方法 43
5.10視覺追蹤之伺服控制 45
5.11目標物奔跑視覺追蹤之伺服控制 46
第六章 結論與未來展望 47
參考文獻 48

[1]Yuan-Hao Lai and Chuan-Kai Yang, ” Video object retrieval by trajectory and appearance,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 25, no. 6,pp.1026-1037, JUNE 2015.
[2]Luis Merino, Fernando Caballero, J.R. Martínez-de Dios and Aníbal Ollero,” Cooperative Fire Detection using Unmanned Aerial Vehicles”, Proceedings of the IEEE International Conference on Robotics and Automation,pp.1884-1889, April,2005.
[3]Dae-Eon Kim and Dong-Soo Kwon, ”Pedestrian detection and tracking in thermal images using shape features,” IEEE 12th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI 2015), Goyang city, Korea,pp.22-25, October 28-30, 2015.
[4]Jan Portmann, Simon Lynen, Margarita Chli and Roland Siegwart,”People detection and tracking from aerial thermal views,” IEEE International Conference on Robotics & Automation (ICRA), Hong Kong, China,pp.1794-1880, May 31 - June 7,2014.
[5]D. Comaniciu, V. R. Member, and P. Meer, ”Kernel-based object tracking,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 25, no. 5,pp. 564–575, May 2003.
[6]David Gibson and Michael Spann, ”Robust Optical Flow Estimation Based on a Sparse Motion Trajectory Set,” IEEE Transactions on Image Processing, vol.12, no.4,pp.431-445, APRIL,2003.
[7]Zalili Binti MUSA, Junzo W A T ADA and Huiming Zhang,” Multi-Camera Tracking Mehtod Based on Particle Filtering”, World Automation Congress (WAC), pp.1-8,19-23 Sept,2010.
[8]C. M. Huang, Y. R. Chen, and L. C. Fu, ”Visual tracking of human head and arms using adaptive multiple importance sampling on a single camera in cluttered environments,” IEEE Sensors Journal, vol. 14, no. 7, pp.2267-2275, 2014.
[9]H. Cho, P. E. Rybski and W. Zhang “Vision-based 3D bicycle tracking using deformable part model and interacting multiple model filter,” IEEE International Conference on Robotics and Automation Shanghai International Conference, Shanghai, China, pp.4391-4398, May 9-13, 2011.
[10]W. Zhong, H. Lu, M. H. Yang, ”Robust object tracking via sparse collaborative appearance model,” IEEE Transactions on Image Processing, pp.2356-2368, MAY, 2014.
[11]J. Gong,Y. Sun and Y. Jiang, “A method of multi-target detecting and velocity estimation using automotive radar,” IEEE 17th International Conference on Intelligent Transportation Systems (ITSC), Qingdao, China, pp.2930-2935, October 8-11, 2014.
[12]O. Kermorgant and F. Chaumette, ”Multi-sensor data fusion in sensor-based control: application to multi-camera visual servoing,” IEEE International Conference on Robotics and Automation Shanghai International Conference Center, Shanghai, China, pp.4518-4523, May 9-13, 2011.
[13]A. Assa and F. J. Sharifi, ”Decentralized multi-camera fusion for robust and accurate pose estimation,” IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), Wollongong, Australia, pp.1696-1701, July 9-12, 2013.
[14]Wai Kit Wong, Zeh Yang Chew, Chu Kiong Loo, Way Soong Lim, “An effective trespasser detection system using thermal camera,” IEEE International Conference on Computer Research and Development, Kuala Lumpur, pp.702-706, May 7-10, 2010.
[15]M. Talha and R. Stolkin, ”Particle filter tracking of camouflaged targets by adaptive fusion of thermal and visible spectra camera data,” IEEE Sensors Journal, vol.14, no.1, pp.159-166, 2014.
[16]R. Stolkin, D. Rees, M. Talha, and I. Florescu, ”Bayesian fusion of thermal and visible spectra camera data for region based tracking with rapid background adaptation,” IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), Hamburg, Germany, pp.192-199, September 13-15, 2012.
[17]J. Heo, S. G. Kong, B. R. Abidi, and M. A. Abidi, “Fusion of visual and thermal signatures with eyeglass removal for robust face recognition,”in Proc. Workshop OTCBVS, Washington, DC, USA, p.122, Jul. 2004.
[18]Olivier Barnich and Marc Van Droogenbroeck,” IEEE Transactions on image processing, vol. 20, no. 6, pp.1709-1724, JUNE 2011.
[19]Ismail Ben Ayed, Kumaradevan Punithakumar, and Shuo Li, ” Distribution matching with the bhattacharyya similarity: A Bound Optimization Framework,” IEEE Transactions on pattern analysis and machine intelligence, vol. 37, no. 9,pp.1777-1791, SEPTEMBER 2015.
[20]A. Y. Yang, S. S. Sastry, A. Ganesh, Y. Ma, ”Fast l1-minimization algorithms and an application in robust face recognition: a review,” IEEE 17th International Conference on Image Processing, Hong Kong, pp.1849-1852, September 26-29, 2010.