本研究建立一個以 FPGA 實現移動相機對移動目標之實時偵測系統。影像處理程序方面，將影像裁切為多個區塊後，對連續影像使用樣板比對法以建立各比對區塊之位移。接著，透過隨機取樣一致算法對由各比對區塊之位移所建立的仿射變換矩陣參數進行評價以找出最好的仿射變換模型，完成相機移動估測。其中，以線性反饋移位暫存器產生隨機取樣一致算法所需的隨機數。將連續影像之背景對齊後，使用幀間差分法將背景去除並留下移動目標。最後，利用自適應二值化方法來凸顯移動目標，於影像中標記出移動目標，完成移動目標偵測。為確保硬體電路滿足邏輯上、時序上的要求，並可在一幀影像的時間內完成處理程序，研究中相機取樣頻率為 26 Hz ，影像處理程序時脈為 12.5 MHz 。實驗結果顯示，本論文建議之方法可完成移動相機對移動目標之實時偵測。

In this study, a real-time image processing system based on FPGA for detecting moving targets with a moving camera is established. For image processing, any two successive images are segmented into multiple blocks and the relative displacement of each block is estimated with block matching method. Then, Random Sample Consensus (RANSAC) is utilized to evaluate the best affine transform model built from the result of block matching method and the motion of a camera can be estimated. In particular, Linear Feedback Shift Register (LFSR) is used to generate a pseudo random sequence for RANSAC. After registering successive images, the background is removed by frame differencing method. Finally, moving targets are distinguished and marked with adaptive threshold method to complete the moving target detection. In order to fulfill logic and timing requirements in the hardware circuit and guarantee the image processing can be completed within one frame sampling period, the frame rate is set at 26 Hz and the clock of image processing is set at 12.5 MHz. According to the experiment result, moving targets captured with a moving camera can be detected in real-time with the proposed method.

[1] RANSAC wiki. In https://reurl.cc/KAjRYq.
[2] Cheng Xing and J.J. Huang. An improved mosaic method based on sift algorithm for uav sequence images. In 2010 International Conference On Computer Design and Applications, 2010.
[3] J.I. Chen and J.T. Chang. Applying a 6-axis mechanical arm combine with computer vision to the research of object recognition in plane inspection. Journal of Articial Intelligence, 2(02):77-99, 2020.
[4] Chang, Shyang-Lihet al. Automatic license plate recognition. IEEE transactions on intelligent transportation systems, 5(1):42-53, 2004.
[5] Stepan Komkov and Aleksandr Petiushko. Advhat: Real-world adversarial attack on arcface face id system. In 2020 25th International Conference on Pattern Recognition (ICPR), pages 819-826. IEEE, 2021.
[6] Mahmood Fathy and Mohammed Yakoob Siyal. An image detection technique based on morphological edge detection and background dierencing for real-time traffic analysis. Pattern Recognition Letters, 16(12):1321-1330, 1995.
[7] Chanho Lee and J.H. Moon. Robust lane detection and tracking for real-time applications. IEEE Transactions on Intelligent Transportation Systems,19(12):4043-4048, 2018.
[8] Ahmed, Afzal et al. Uav based monitoring system and object detection technique development for a disaster area. International Archives of Photogrammetry, Remote Sensing and Spatial information Sciences, 37:373-377, 2008.
[9] N. Yamamoto and N. Uchida. Improvement of image processing for a collaborative security
flight control system with multiple drones. In 2018 32nd International Conference on Advanced Information Networking and Applications Workshops (WAINA), pages 199-202, 2018.
[10] Lopez-Bravo, A et al. FPGA-based video system for real time moving object detection. In CONIELECOMP 2013, 23rd International Conference on Electronics, Communications and Computing, pages 92-97. IEEE,2013.
[11] Barbara Zitova and Jan Flusser. Image registration methods a survey. Image and vision computing, 21(11):977-1000, 2003.
[12] J Alison Noble. Finding corners. Image and vision computing, 6(2):121-128,1988.
[13] Maes, F. et al. Medical image registration using mutual information. Proceedings of the IEEE, 91(10):1699-1722, 2003.
[14] Min Xu and Pramod K Varshney. A subspace method for fourier-based image registration. IEEE Geoscience and Remote Sensing Letters, 6(3):491-494,2009.
[15] Daniel I Barnea and Harvey F Silverman. A class of algorithms for fast digital image registration. IEEE transactions on Computers, 100(2):179-186, 1972.
[16] P. Anandan. Measuring visual motion from image sequences. University of Massachusetts Amherst, 1987.
[17] Di Stefano, Luigi et al. An ecient algorithm for exhaustive template matching based on normalized cross correlation. In 12th International Conference on Image Analysis and Processing, 2003. Proceedings., pages 322-327. IEEE, 2003.
[18] Hisham, MB et al. Template matching using sum of squared dierence and normalized cross correlation. In 2015 IEEE student conference on research and development (SCOReD), pages 100-104. IEEE, 2015.
[19] R.C. Gonzalez and R.E. Woods. Digital Image Processing. Upper Saddle River, 2008.
[20] J.W. Tang et al. FPGA-based real-time moving target detection system for unmanned aerial vehicle application. International Journal of Recongurable Computing, 2016.
[21] C.H. Li. A study on the recognition and acquisition of sports highlight-moment images. Master's thesis, National Taiwan Ocean University, 2018.
[22] N. Otsu. A threshold selection method from gray-level histograms. IEEE Trans. Sys., Man., 9(1):62-66, 1979.
[23] LUC DE VOS and MICHAEL STEGHERR. Parameterizable vlsi architectures for the full- search block-matching algorithm. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 36(10), 1989.
[24] Synchronous Pipelined, single cycle deselect static RAM, 2007.
[25] Raj S Katti and Sudarshan K Srinivasan. Ecient hardware implementation of a new pseudo-random bit sequence generator. In 2009 IEEE International Symposium on Circuits and Systems, pages 1393-1396. IEEE, 2009.
[26] Alhadawi, Hussam S et al. Designing a pseudorandom bit generator based on lfsrs and a discrete chaotic map. Cryptologia, 43(3):190-211, 2019.
[27] Mark Goresky and Andrew M Klapper. Fibonacci and galois representations of feedback-with-carry shift registers. IEEE Transactions on Information Theory, 48(11):2826-2836, 2002.
[28] PMI KM33-05A-400N0-0331. http://www.nsk-cs.com.tw/?f=km-monostage.
[29] TRDB-D5M Hardware specication V0.2, 2009.
[30] DE2 70 User manual v105, 2008.
[31] Thomas Komarek and Peter Pirsch. Array architectures for block matching algorithms. IEEE Transactions on Circuits and Systems, 36(10):1301-1308,1989.