本論文提出一套無人機夜間無人機追蹤識別系統，此系統以RGB影像為輸入，若檢測到可疑物件會回傳至使用者端提醒使用者，必要時可對特定物件進行追蹤，用以輔佐特定場域之全天監視、海岸線偷渡與走私巡檢等使用期間會涵蓋到夜間且巡視範圍較大的任務，可在人力監測的同時多加一份保險，減少因人為因素所產生的失誤。系統將輸入的夜間影像經夜間影像增強算法後，使其亮度盡可能地與日間影像亮度相似，同時避免因過度曝光等導致的失真，讓多數基於日間影像資料作為訓練的物件辨識、目標追蹤演算法，能在輸入經系統強化後的夜間影像有這些演算法原定的效果，以達成夜間運作之可行性，進而實現全天的應用。且為了防止系統無法用於幾乎無光的環境之中如夜間海上，在無人機上搭載探照燈，能提供最低限度的照明，進而保障系統的全面性。系統擁有自動化功能可自行判斷使否進行夜間影像強化，也可在發現可疑目標時能搶佔先機先一步進行追縱並同步通知使用者進行後續關注與判斷。在數據方面配有探照燈時開啟夜間影像增強能有約3~6%不等的提升、而在有環境光下開啟夜間影像增強對於影像中有多目標的辨識率則至少23%左右的提升、而在未配有探照燈並且未有環境光時則無法做辨識。系統的影像處理速率約為15~20FPS，足以應付實時的物件辨識追蹤任務。

This thesis proposes an all-day Unmanned Aerial Vehicle (UAV) surveillance system. With the night vision enhancement model, this system can not only detect suspicious object in daytime but also at night. The system uses only RGB image as input. Based on the environment the night enhancement model can be activated if too dark, otherwise deactivate. The night enhancement model uses a deep learning algorithm to enhance the input image to allow better object detection. The UAV surveillance system can even track the suspicious object. When it equips with a spotlight (whether environment is bright or not) and uses the night enhancement model that will increase object detection rate by about 3~6%. When the environment is bright (whether equipped with a spotlight or not), uses the night enhancement model will increase object detection rate by at least 23%. Frame rate is around 15~20FPS that is feasible for a real-time tracking mission.

論文審定書 i
摘 要 ii
Abstract iii
目 錄 iv
圖 目 錄 vi
表 目 錄 viii
第一章 簡介 1
1.1、論文概述 3
1.2、論文貢獻 4
1.3、論文架構 4
第二章 文獻探討 6
2.1、RGB夜間影像增強 6
2.1.1、傳統RGB夜間影像增強 6
2.1.2、基於機器學習之RGB夜間影像增強 9
2.2、熱成像夜間影像增強 12
2.3、無人機使用夜間影像增強 13
2.3.1、使用熱成像 13
2.3.2、使用RGB影像 14
第三章 研究方法 18
3.1、系統整體架構 18
3.2、夜間影像增強 21
3.3、探照設備使用 23
3.4、自動化夜間影像增強 26
3.5、自動化目標追蹤 27
第四章 實驗結果 28
4.1、地面水平視角與無人機視角夜間影像增強結果 29
4.2、地面水平視角、無人機視角與環境光照差異比較 30
4.3、實機夜間飛行測試 33
4.3.1、光照充足時使用結果 34
4.4、探照設備使用結果 35
4.4.1、不同高度比較 36
4.4.2、不同環境光比較 39
4.5、物件辨識使用夜間影像增強與否結果對比 40
4.6、實機夜間追蹤 44
4.7、自動化夜間影像增強 46
4.8、自動化目標追蹤 47
第五章 結論與未來工作 48
5.1 結論 48
5.2 未來工作 48
參考文獻 50

[1] J. Ye, C. Fu, G. Zheng, Z. Cao, and B. Li, "DarkLighter: Light up the darkness for UAV tracking," in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021: IEEE, pp. 3079-3085.
[2] A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "Yolov4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020.
[3] B. Li, W. Wu, Q. Wang, F. Zhang, J. Xing, and J. Yan, "Siamrpn++: Evolution of siamese visual tracking with very deep networks," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4282-4291.
[4] I. Goodfellow et al., "Generative adversarial networks," Communications of the ACM, vol. 63, no. 11, pp. 139-144, 2020.
[5] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, 1998.
[6] S. M. Pizer et al., "Adaptive histogram equalization and its variations," Computer vision, graphics, and image processing, vol. 39, no. 3, pp. 355-368, 1987.
[7] E. H. Land, "The retinex theory of color vision," Scientific american, vol. 237, no. 6, pp. 108-129, 1977.
[8] E. H. Land, "Recent advances in retinex theory," in Central and Peripheral Mechanisms of Colour Vision: Proceedings of an International Symposium Held at The Wenner-Gren Center Stockholm, June 14–15, 1984, 1985: Springer, pp. 5-17.
[9] D. J. Jobson, Z.-u. Rahman, and G. A. Woodell, "Properties and performance of a center/surround retinex," IEEE transactions on image processing, vol. 6, no. 3, pp. 451-462, 1997.
[10] G. Hines, Z.-u. Rahman, D. Jobson, and G. Woodell, "Single-scale retinex using digital signal processors," in Global signal processing conference, 2005, no. Paper 1324.
[11] D. J. Jobson, Z.-u. Rahman, and G. A. Woodell, "A multiscale retinex for bridging the gap between color images and the human observation of scenes," IEEE Transactions on Image processing, vol. 6, no. 7, pp. 965-976, 1997.
[12] Z.-u. Rahman, D. J. Jobson, and G. A. Woodell, "Multi-scale retinex for color image enhancement," in Proceedings of 3rd IEEE international conference on image processing, 1996, vol. 3: IEEE, pp. 1003-1006.
[13] X. Yang et al., "Implementing real-time RCF-Retinex image enhancement method using CUDA," Journal of Real-Time Image Processing, vol. 16, pp. 115-125, 2019.
[14] X. Guo, Y. Li, and H. Ling, "LIME: Low-light image enhancement via illumination map estimation," IEEE Transactions on image processing, vol. 26, no. 2, pp. 982-993, 2016.
[15] X. Dong, Y. Pang, and J. Wen, "Fast efficient algorithm for enhancement of low lighting video," in ACM SIGGRAPH 2010 Posters, 2010, pp. 1-1.
[16] Z. Ying, G. Li, Y. Ren, R. Wang, and W. Wang, "A new low-light image enhancement algorithm using camera response model," in Proceedings of the IEEE international conference on computer vision workshops, 2017, pp. 3015-3022.
[17] Y. Jiang et al., "Enlightengan: Deep light enhancement without paired supervision," IEEE transactions on image processing, vol. 30, pp. 2340-2349, 2021.
[18] Y. Meng, D. Kong, Z. Zhu, and Y. Zhao, "From night to day: GANs based low quality image enhancement," Neural Processing Letters, vol. 50, pp. 799-814, 2019.
[19] P. Alaspure, P. Hambarde, A. Dudhane, and S. Murala, "DarkGAN: night image enhancement using generative adversarial networks," in Computer Vision and Image Processing: 5th International Conference, CVIP 2020, Prayagraj, India, December 4-6, 2020, Revised Selected Papers, Part I 5, 2021: Springer, pp. 293-302.
[20] C. Chen, Q. Chen, J. Xu, and V. Koltun, "Learning to see in the dark," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3291-3300.
[21] C. Wei, W. Wang, W. Yang, and J. Liu, "Deep retinex decomposition for low-light enhancement," arXiv preprint arXiv:1808.04560, 2018.
[22] W. Yang, W. Wang, H. Huang, S. Wang, and J. Liu, "Sparse gradient regularized deep retinex network for robust low-light image enhancement," IEEE Transactions on Image Processing, vol. 30, pp. 2072-2086, 2021.
[23] W. Yang, S. Wang, Y. Fang, Y. Wang, and J. Liu, "From fidelity to perceptual quality: A semi-supervised approach for low-light image enhancement," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 3063-3072.
[24] G. Li, Y. Yang, X. Qu, D. Cao, and K. Li, "A deep learning based image enhancement approach for autonomous driving at night," Knowledge-Based Systems, vol. 213, p. 106617, 2021.
[25] Y. Choi, N. Kim, S. Hwang, and I. S. Kweon, "Thermal image enhancement using convolutional neural network," in 2016 IEEE/RSJ international conference on intelligent robots and systems (IROS), 2016: IEEE, pp. 223-230.
[26] C. Li, W. Xia, Y. Yan, B. Luo, and J. Tang, "Segmenting objects in day and night: Edge-conditioned CNN for thermal image semantic segmentation," IEEE Transactions on Neural Networks and Learning Systems, vol. 32, no. 7, pp. 3069-3082, 2020.
[27] A. Bhandari, A. Kafle, P. Dhakal, P. R. Joshi, and D. B. Kshatri, "Image enhancement and object recognition for night vision surveillance," arXiv preprint arXiv:2006.05787, 2020.
[28] onyxstar. "FOX-C8 XT." https://www.onyxstar.net/fox-c8-xt/ (accessed.
[29] T. Müller and B. Erdnüß, "Robust drone detection with static VIS and SWIR cameras for day and night counter-UAV," in Counterterrorism, Crime Fighting, Forensics, and Surveillance Technologies III, 2019, vol. 11166: SPIE, pp. 58-72.
[30] T. Müller, "Robust drone detection for day/night counter-UAV with static VIS and SWIR cameras," in Ground/Air Multisensor Interoperability, Integration, and Networking for Persistent ISR VIII, 2017, vol. 10190: SPIE, pp. 302-313.
[31] W. Wang, Y. Peng, G. Cao, X. Guo, and N. Kwok, "Low-illumination image enhancement for night-time UAV pedestrian detection," IEEE Transactions on Industrial Informatics, vol. 17, no. 8, pp. 5208-5217, 2020.
[32] B. Li, C. Fu, F. Ding, J. Ye, and F. Lin, "ADTrack: Target-aware dual filter learning for real-time anti-dark UAV tracking," in 2021 IEEE international conference on robotics and automation (ICRA), 2021: IEEE, pp. 496-502.
[33] B. Li, C. Fu, F. Ding, J. Ye, and F. Lin, "All-day object tracking for unmanned aerial vehicle," IEEE Transactions on Mobile Computing, 2022.
[34] J. Zhu et al., "DCPT: Darkness Clue-Prompted Tracking in Nighttime UAVs," arXiv preprint arXiv:2309.10491, 2023.