生命科學的研究在近年越來越受重視，藉由研究果蠅之類的昆蟲體內較為單純的大腦與各個神經元之間傳導物質和路徑的相互作用關係，其最終目的是希望有朝一日也能夠對人類大腦中的各種運作機制有更深入明確的了解。在這些研究中，都必須依靠大量的共軛焦顯微鏡影像資料來協助。這些影像中，也會因為不同的要求而需要經過某些影像處理。由於共軛焦顯微鏡的取像解析度大小是有極限的，因此取得的部份影像集合會因為取像的過程而產生許多缺陷。

這篇論文提出了一種縫合影像的系統，主要是用來縫合這些部份影像的集合。我們以金字塔（pyramid）架構為基礎實行對影像的多重解析度縫合（multi-resolution blending）；同時做了一些修改以符合這種特殊的影像狀況。另外，這個系統利用次像素位移（sub-pixel displacement）的分析將影像的縫合和對位結合在一起對影像做處理：對位結果會影響縫合結果，同時縫合結果也會影響對位結果。利用這個機制，我們可以重複地對對位以及縫合做調整直到結果影像最佳化。而透過實驗，可以知道我們提出的方法對於目標影像處理會得到不錯的結果。這些結果證實了我們對這種特殊影像集合的假設跟預想沒有問題，最終得到的縫合結果影像則會能夠保留最多的原始影像資訊，同時維持一定的平滑度。

The research in the area of Life Science has aroused more and more attention recently. By studying on the simpler neural interactions, which are established by neurotransmitters and their transition pathways inside the Drosophila brains, the ultimate goal is to figure out the neurons’ working mechanism in the Human Brain. In these researches, lots of confocal microscope images are acquired first, and then a series of image processing operations are applied to them so that these images would become suitable to further analysis by biologists. Besides, since the resolution of acquired images from confocal microscope has its limitation, a whole brain structure may be imaged into several partial image sets, and therefore, the image acquisition process may lead to some defects in the images.

In this study, we propose an image stitching system for dealing with the partial image sets mentioned above. Based on pyramid structure, we make some changes to fit this particular kind of images and then accomplish multi-resolution blending. Moreover, this system also uses sub-pixel displacement analysis such that the image stitching method and registration algorithm – the results of these two may affect each other – can be integrated. By such mechanism, we can recursively adjust stitching position (registration) and blending coefficients (stitching) until the resultant image achieves an optimum. Through experimental results, it is shown that the proposed method would provide good results for target images. These results also confirm our assumption and presupposition for this particular kind of image set. The stitched image can keep as much information of original ones as possible and also maintain a certain level of smoothness.

1.PETER J. BURT and EDWARD H. ADELSON,　“A Multiresolution Spline With Application to Image Mosaics”,　ACM Transactions on Graphics, Vol. 2, No. 4, pp.217-236, Oct. 1983

2.PETER J. BURT,　“Fast filter transforms for image processing”,　Computer Vision. Graphics and Image Process. Vol. 16, pp.20-51, 1981

3.PETER J. BURT and EDWARD H. ADELSON,　“The Laplacian pyramid as a compact image code”,　IEEE Trans. Communications. Vol. COM-31, pp.532-540, Apr. 1983

4.S.L. TANIMOTO, and T. PAVLIDIS,　“A hierachical data structure for picture processing”,　Computer. Graphics and Image Processing. Vol. 4, pp. 104-119, Jun. 1975

5.D. MARR, and E. HILDRETH,　“Theory of edge detection”,　Proceedings of the Royal Society, London, Vol. B-207, 1980, pp.187-217.

6.Barbara Zitová, and Jan Flusser,　“Image registration methods: a survey”,　Image and Vision Computing, Vol.21, pp.977–1000, 2003

7.Josien P. W. Pluim, J. B. Antoine Maintz and Max A. Viergever,　“Mutual information based registration of medical images: a survey”,　IEEE Transactions on Medical Imaging, Volume 22, Issue 8, pp.986–1004, Aug. 2003

8.D. L. G. Hill, D. J. Hawkes, N. A. Harrison and C. F. Ruff,　“A strategy for automated multimodality image registration incorporating anatomical knowledge and imager characteristics”, in Information Processing in Medical Imaging, H. H. Barrett and A. F. Gmitro, Eds. 1993, vol. 687 of Lecture Notes in Computer Science, pp.182–196, Springer-Verlag, Berlin

9.Jiaya Jia, Chi-Keung Tang,　“Eliminating Structure and Intensity Misalignment in Image Stitching”,　Proceedings of the Tenth IEEE International Conference on Computer Vision, October 17-20, 2005, pp.1651-1658,

10.A. Zomet, A. Levin, S. Peleg, and Y. Weiss,　“Seamless image stitching by minimizing false edges”,　IEEE Trans. Image Processing, Vol. 15, Issue 8, pp. 969-977, 2005