本篇論文針對智慧型相機提供一個新穎的功能—將不必要的物體移除。在先前相關的研究中，紋理合成和Image inpainting建構了影像補償未知區域的基礎知識。紋理合成的方法可以將大的未知區域填入輸入圖檔的紋理，但大量的運算時間是很嚴重的缺點。而Image inpainting可以修復影像的刮痕，但在處理大的未知區域時會有影像模糊的問題。本篇論文提供了一個新穎的演算法，結合了子區塊紋理合成技術和加權內插法來克服上述兩個主要方法的缺點。

　　首先，我們對輸入圖檔執行色彩紋理分佈分析來決定不同種類紋理時適合的方法。當在同質紋理時，加權內插法可以減少大量的運算時間。而在處理不同質紋理時，子區塊紋理合成在比對公式中加入拉普拉斯梯度來準確的連接邊界。此外，我們提供了一個根據克許邊界運算子和色彩比率梯度的缺陷偵測機制。這個缺陷偵測機制能在沒有人為介入下修補不合理的部分。我們用傳統的紅綠藍色彩空間來得到相當優秀的輸出影像。除此之外，我們演算法的規則性比其他先前提出的演算法更適合來做硬體的實現。

　This work aims for a novel function for smart camera—redundant object removal from digital photograph. In previous related researches, texture synthesis and image inpainting construct the fundamentals of filling the lost region in image. Texture synthesis can be used to fill the large hole of input texture, but the exhaustive computation time is the crucial drawback. Image inpainting can be used to repair the scratches of image, unfortunately it will produce the blurring effect while dealing with large lacuna region. In this paper, we propose a novel object removal algorithm composed of sub-patch texture synthesis technique and weighted interpolation method to overcome the drawback of these two approaches. Firstly, the color texture distribution analysis is performed to decide the proper method in the different type of texture region. The exhaustive computation time is reduced by the weighted interpolation method in the homogeneous texture region. In the inhomogeneous texture region applied sub-patch texture synthesis technique, the gradient of Laplacian is added in the matching formula to connect edges precisely. Besides, we provide an artifact detection mechanism based on the Kirsch edge operator and color ratio gradients to detect the synthesized violation blocks in the target region. The artifact detection mechanism can lead repainting the violation blocks without user intervention. The proposed algorithm achieves better performance with seamless output images by traditional RGB color space. Moreover, the regular computation is more suitable for hardware implementation than previous existing algorithms.

[1]A. Efros and T. Leung, “Texture synthesis by non-parametric sampling,” in Proc. Int. Conf. Computer Vision, Kerkyra, Greece, Sep. 1999, pp. 1033-1038.
[2]L. Y. Wei and M. Levoy, “Fast texture synthesis using tree-structured vector quantization,” in Proc. ACM Conf. Computer Graphics (SIGGRAPH), pp. 479-488, July, 2000.
[3]M. Ashikhmin, “Synthesizing natural textures,” in Proc. ACM Symp. Interactive 3D Graphics, Mar. 2001, pp. 217-226.
[4]S. C. Pei, Y. C. Zeng, and C. H. Chang, “Virtual restoration of ancient Chinese paintings using color contrast enhancement and lacuna texture synthesis,” IEEE Trans. Image Processing, vol. 13, no. 3, pp. 416-429, 2004.
[5]A. Efros and W. T. Freeman, “Image quilting for texture synthesis and transfer,” in Proc. ACM Conf. Computer Graphics (SIGGRAPH), Aug. 2001, pp. 341-346.
[6]L. Liang, C. Liu, Y. Q. Xu, B. Guo, and H. Y. Shum, “Real-time texture synthesis by patch-based sampling”, in ACM Trans. on Graphics, vol. 20, pp. 127-150, 2001.
[7]Q. Wu, and Y. Yu, “Feature matching and deformation for texture synthesis,” in ACM Trans. Graphics (TOG), vol. 23, no. 3, Aug. 2004.
[8]M. Bertalmio, G. Sapiro, V. Caselles, and C. Ballester, “Image inpainting,” in Proc. ACM Conf. Computer Graphics (SIGGRAPH), July. 2000, pp. 417-424.
[9]M. Bertalmio, L. Vese, G. Sapiro, and S. Osher, “Simultaneous structure and texture image inpainting,” IEEE Trans. Image Processing, vol. 12, no. 8, August, 2003.
[10]A. Criminisi, P. P’erez, and K. Toyama, “Region filling and object removal by examplar-based image inpainting,” IEEE Trans. Image Processing, vol. 13, Sep. 2004.
[11]I. Drori, D. Cohen-Or, H. Yeshurun, “Fragment-based image completion,”in ACM Trans. Graphics (SIGGRAPH), vol.22, no. 3, pp. 303-312, 2003.
[12]J. Jia, and C. K. Tang, “Inference of segmented color and texture description by tensor voting,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 26, June, 2004.
[13]J. F. Wang, H. J. Hsu, and S. C. Liao, “A novel framework for object removal from digital photogragh,” in Proc. IEEE Conf. Image Processing, 2005. (accepted)
[14]R. C. Gonzalez and R. E Woods, Digital Image Processing. Norwell, MA: Addison-Wesley, 1992.
[15]P. Ndjiki-Nya, M. Kootz, and T. Wiegand, “Automatic detection of video synthesis related artifacts,” in Proc. IEEE Conf. Acoustics, Speech, and Signal Processing, 2004.
[16]T. Gevers, "Image Segmentation and Similarity of Color-Texture Objects," IEEE Tran. Multimedia, vol.4, no. 4, 2000.
[17]R. Bornard, E. Lecan, L. Laborelli, and J. H. Chenot, “Missing data correction in still images and image sequences,” in ACM Multimedia, Dec. 2002.