在影像處理的領域中，物件辨識一直是一個重要的課題。而在物件辨識中的方法設計上，會探討此方法是否能保有旋轉不變性、位移不變性、縮放不變性等議題，但在不同的應用中並不一定會全部包含上述三種不變性；在自動化光學檢測上，如工業品質檢測，此應用是希望能得到物件辨識及其對於物件旋轉角度之估計，所以對於上述之不變性較注重的是在於旋轉不變性上，且對於高精確度以及即時性有相當高之要求，此即為需要能在限制時間內達到最高之精確度。在此篇論文中針對此問題，提出了一個新穎之物件辨識特徵描述器及其演算法，此描述器是基於物件輪廓作為特徵點，並能精確之求出其物件相識度及其旋轉角度，並能順帶得到輪廓瑕疵區域；而使用物件輪廓的好處是，相較基於像素點之描述器能更加明確描述物件，且較能不受外界之影響，如遮蔽、光源等，對於精確度之問題上能有所提升。另外在坐標系之使用上，此篇論文使用了極座標系，極座標系的優點相較於笛卡兒座標系，能提供像素點之長度及角度資訊，對於特徵點擷取的設計上助益很大。

Object recognition is an important issue in image processing. For different object recognition methods, we may discuss whether the method includes rotation, scaling, or translation invariances or not. However, not every application involves all three invariances. In Automated Optical Inspection (ex: Industrial quality testing) for example, an application would hope to obtain an estimate on the objects rotation angle, and thus has to focus on the rotation variant above the other invariances. The application also requires high accuracy and low time complexity to achieve the goal. To solve this problem, this paper proposes a novel shape descriptor and the associated algorithm. The descriptor is contour-based and uses the contour points as an object’s feature. The goal is to acquire the similarity between two objects and the difference between their respective rotation angles. It can also detect defective areas on the object’s contour. The advantage of using contour-based detection as opposed to using a region-based method is that it is able to describe the object’s information more clearly and can reduce outside influences like occlusion, brightness and so on. With the above advantages, the accuracy can also be improved. Additionally, the proposed will utilize the Polar coordinate system instead of the Cartesian coordinate system. The benefit of using said coordinate system is that a pixel’s radius and angle can now be provided as information, which is very helpful when capturing the object’s feature.

[1] L. d. F. D. Costa and R. M. C. Jr, "Shape Analysis and Classification: Theory and Practice," CRC Press, 2000.
[2] S. Loncaric, "A survey of shape analysis techniques," Pattern Recognition 31, pp. 983-1001, 1998.
[3] C. Xu, J. Liu and X. Tang, "2D Shape Matching by Contour Flexibility," IEEE Trans. Pattern Analysis and Machine Intelligence 31, pp. 180-186, 2009.
[4] W.-Y. Kim and Y.-S. Kim, "A region-based shape descriptor using Zernike moments," Signal Process. Image Commun 16, pp. 95-102, 2000.
[5] S. X. Liao and M. Pawlak, "On image analysis by moments," IEEE Trans. Pattern Analysis and Machine Intelligence 18, pp. 254-266, 1996.
[6] D. Zhang and G. Lu, "Shape-Based Image Retrieval Using Generic Fourier Descriptor," Signal Processing: Image Communication 17, pp. 825-848, 2002.
[7] A. Khotanzad and Y. H. Hong, "Invariant image recognition by zernike moments," IEEE Trans. Pattern Anal. Mach.Intell., vol. 12, no. 5, pp., pp. 489-497, 1990.
[8] C. Kan and M. D. Srinath, "Invariant character recognition with zernike and orthogonal fourier-mellin moments," vol. 35, no. 1, pp., p. 143–154, 2002.
[9] W.-Y. Kim and Y.-S. Kim, "Robust rotation angle estimator," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 768 - 773, 1999.
[10] J. Revaud, G. Lavoué and A. Baskurt, "Improving Zernike Moments Comparison for Optimal Similarity and Rotation Angle Retrieval," IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, pp. 627 - 636, 2008.
[11] F. Mokhtarian, S. Abbasi and J. Kittler., "Efficient and robust retrieval by shape content through curvature scale space," Image Databases and Multi-Media Search, pp. 51-58, 1997.
[12] Y. Rui, A. C. She and T. S. Huang, "A modified Fourier descriptor for shape matching in MARS," Image Databases and Multimedia Search 8, pp. 165-180, 1998.
[13] I. Bartolini, P. Ciaccia and M. Patella, "M. Patella. WARP: Accurate Retrieval of Shapes Using Phase of Fourier Descriptors and Time Warping Distance," IEEE Trans. Pattern Analysis and Machine Intelligence 27, pp. 142-147, 2005.
[14] X. Shu and Xiao-JunWu, "A novel contour descriptor for 2D shape matching and its application to image retrieval," Image and vision Computing, pp. 286-294, 2011.
[15] E. Levina and P. Bickel, "The EarthMover’s Distance is the Mallows Distance: Some Insights from Statistics," Proceedings of ICCV 2001, p. 251–256, 2001.
[16] M. Anelli, L. Cinque and E. Sangineto, "Deformation tolerant generalized Hough transform for sketch-based image retrieval in complex scenes," Image and Vision Computing 25, pp. 1802-1813, 2007.
[17] H.-K. Kim and J.-D. Kim, "Region-based shape descriptor invariant to rotation, scale and translation," Signal Processing: Image Comm. 16, pp. 87-93, 2000.
[18] J. Canny, "A Computational Approach To Edge Detection," IEEE Trans. Pattern Analysis and Machine Intelligence, 8, p. 679–698, 1986.
[19] S. Belongie, J. Malik and J. Puzicha, "Shape matching and object recognition using shape contexts," IEEE Trans. Pattern Analysis and Machine Intelligence 24, pp. 509-522, 2002.
[20] M. J. Swain and D. H. Ballard, "Color indexing," International Journal of Computer Vision 7, pp. 11-32, 1991.
[21] C. W. Niblack and et al, "Querying images by content, using color, texture, and shape," In SPIE Conference on Storage and Retrieval for Image and Video Databases 1908, pp. 173-187, 1993.
[22] A. Marzal and V. Enrique , "Computation of normalized edit distance and applications," IEEE Trans.Pattern Analysis and Machine Intelligence 15, pp. 926-932, 1993.
[23] C. Gentsos and et al, "Realtime canny edge detection parallel implementation for FPGAs," IEEE ICECS, p. 499–502, 2010.
[24] W. He and K. Yuan, "An improved canny edge detector and its realization on FPGA," IEEE 7th WCICA, p. 6561–6564, 2008.
[25] E. F. Krause, Taxicab Geometry, Dover, 1987.