近年來，由於感光元件的進步，圖像的解析度已超過一億像素，這可能會導致在影像處理階段處理過多資訊。超像素技術在影像處理中可被視為一項前處理技術，此技術可以降低輸入圖的複雜度。一個超像素是由許多有相似顏色或特徵的像素所組成。
現存的超像素演算法有過度切割的問題，這可能會造成簡單結構的圖片卻仍需進行複雜的運算。在這篇論文中，我們提出一個基於顏色差異與區域鄰接圖的迭代超像素合併演算法，此方法運用區域鄰接圖代表超像素，此舉可確保超像素之間的相鄰關係。接下來，我們計算超像素間的顏色差異及合併閥值。最後，我們藉由顏色差異及閥值來合併超像素，迭代地執行這個階段直到沒有超像素可以再被合併。如此一來，我們可以大幅減少冗餘的超像素及改善過度切割的缺點。此演算法可以大幅減少高達50%的冗餘超像素且可維持準確率。

In recent years, owing to the advance of the image sensors, the resolution of images achieves more than one hundred megapixels, which lead to numerous information for image processing. Thus, the superpixel technique can be regarded as a pre-processing step of image processing for reducing the complexity of the input image.
In the existing superpixel algorithms, a common shortcoming is over segmentation which may cause complex computation even for simple structured images. In this thesis, we propose an iterative superpixel merging algorithm based on color differences and the region adjacent graph. The proposed method uses region adjacent graph to represent superpixels, which exactly confirms the adjacent relationship between superpixels. Then, we compute the color differences between superpixels and define the merging threshold. Finally, we merge superpixels according to color differences and merging threshold, and we iteratively perform this step to all superpixels and its adjacent superpixels until no more superpixels can be merged. By doing so, we can greatly reduce the redundancy superpixels and improve the shortcoming of the over segmentation. The proposed method can effectively decrease up to 50% redundancy superpixels and still maintain the accuracy.

[1] Z. Li, X.-M. Wu, and S.-F. Chang, "Segmentation using superpixels: A bipartite graph partitioning approach," in 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012: IEEE, pp. 789-796.
[2] Y. Li, J. Sun, C.-K. Tang, and H.-Y. Shum, "Lazy snapping," presented at the ACM (SIGGRAPH) 2004.
[3] J.-J. Ding, C.-J. Lin, I.-F. Lu, and Y.-H. Cheng, "Real-time interactive image segmentation using improved superpixels," in 2015 IEEE International Conference on Digital Signal Processing (DSP), 2015: IEEE, pp. 740-744.
[4] C. L. Zitnick and S. B. Kang, "Stereo for image-based rendering using image over-segmentation," International Journal of Computer Vision, vol. 75, no. 1, pp. 49-65, 2007.
[5] S. Iizuka, Y. Endo, Y. Kanamori, J. Mitani, and Y. Fukui, "Efficient depth propagation for constructing a layered depth image from a single image," in Computer Graphics Forum, 2014, vol. 33, no. 7: Wiley Online Library, pp. 279-288.
[6] D. Xiang, T. Tang, L. Zhao, and Y. Su, "Superpixel Generating Algorithm Based on Pixel Intensity and Location Similarity for SAR Image Classification," IEEE Geoscience and Remote Sensing Letters, vol. 10, no. 6, pp. 1414-1418, 2013, doi: 10.1109/LGRS.2013.2259214.
[7] I. Giosan and S. Nedevschi, "Superpixel-based obstacle segmentation from dense stereo urban traffic scenarios using intensity, depth and optical flow information," in IEEE Conference on Intelligent Transportation Systems (ITSC), 8-11 Oct. 2014, pp. 1662-1668, doi: 10.1109/ITSC.2014.6957932.
[8] B. Fulkerson, A. Vedaldi, and S. Soatto, "Class segmentation and object localization with superpixel neighborhoods," in 2009 IEEE 12th international conference on computer vision, 2009: IEEE, pp. 670-677.
[9] R. Achanta, A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Süsstrunk, "SLIC superpixels compared to state-of-the-art superpixel methods," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 34, no. 11, pp. 2274-2282, 2012, doi: 10.1109/TPAMI.2012.120.
[10] A. P. Moore, S. J. Prince, J. Warrell, U. Mohammed, and G. Jones, "Superpixel lattices," in 2008 IEEE conference on computer vision and pattern recognition, 2008: Citeseer, pp. 1-8.
[11] P. Siva and A. Wong, "Grid seams: A fast superpixel algorithm for real-time applications," in 2014 Canadian Conference on Computer and Robot Vision, 2014: IEEE, pp. 127-134.
[12] J.-L. Lin, "A real-time superpixel algorithm based on two-layered grid seams," Master Thesis, Department of Electronic and Computer Engineering, National Taiwan University of Science and Technology,2015.
[13] X. Ren and J. Malik, "Learning a classification model for segmentation," Proceedings Ninth IEEE International Conference on Computer Vision, Nice, France, 2003, pp. 10-17 vol.1, doi: 10.1109/ICCV.2003.1238308
[14] K.-C. Hung and C.-H. Lin, "Iterative Superpixel Segmentation Based on Color Differences," in 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE), 2018: IEEE, pp. 270-271.
[15] C.-Y. Hsu and J.-J. Ding, "Efficient image segmentation algorithm using SLIC superpixels and boundary-focused region merging," in 2013 9th International Conference on Information, Communications & Signal Processing, 2013: IEEE, pp. 1-5.
[16] P. F. Felzenszwalb and D. P. Huttenlocher, "Efficient graph-based image segmentation," International journal of computer vision, vol. 59, no. 2, pp. 167-181, 2004.
[17] K. Haris, S. N. Efstratiadis, N. Maglaveras, and A. K. Katsaggelos, "Hybrid image segmentation using watersheds and fast region merging," IEEE Transactions on image processing, vol. 7, no. 12, pp. 1684-1699, 1998.
[18] ITU-R Recommendation BT.709, "International Telecommunication Union - Radiocommunication sector," 1990.
[19] R. W. G. Hunt and M. R. Pointer, "Measuring Colour," ISBN:9781119975373, 2011.
[20] D. Martin, C. Fowlkes, D. Tal, and J. Malik, "A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics," in Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001, 2001, vol. 2: IEEE, pp. 416-423.
[21] M. Wälivaara, "General Object Detection using Superpixel Preprocessing," Master Thesis, Department of Electrical Engineering, Linköping University, 2017.
[22] Z. Zhang, P. Cui, and W. Zhu, "Deep Learning on Graphs: A Survey," in IEEE Transactions on Knowledge and Data Engineering, doi: 10.1109/TKDE.2020.2981333.