臺灣博碩士論文加值系統

為了獲得更好的生物組織影像，顯微鏡取像在近年來的生科研究中占有重要地位。其中共軛焦顯微鏡被諸多科學家採用來獲取清楚且高解析度的影像。
為了研究果蠅腦的活動機制，腦科學中心利用共軛焦顯微鏡來取得果蠅腦部的影像切片，期望能夠建構出一個易於觀察的立體腦部模型。但是在顯微鏡本身取像有所限制的前提下，每張影像切片間距無法取到與平面解析度一致，這樣取出的物體影像切片在垂直與平行鏡頭方向有著解析度不同的缺點，而建構出的立體影像模型在某些位置上觀察就會顯得模糊，此現象在Z軸方向上最為明顯。
我們設法藉由軟體處理來解決共軛焦顯微鏡天生解析度不同的情形。提出的處理方法要點在於利用顯微鏡對物體攝取多次影像切片，且每次物體本身會調整適當位置。如此同時結合多組影像資訊進行影像內插，一方面在取內插點有多組影像資訊，能夠比以往的線性內插方法來得近；另一方面是結合不同位置掃描的影像切片，內插之後的資訊相較於單一組的影像資料來得豐富。相信藉此能夠達到優於傳統線性內插的結果，便於腦科學中心的科學家來分析並獲得多於以往的生物資訊。
此篇論文的內插方法主要是先替空間的資料點建立分類，之後尋找離內插位置最近的點。利用此方法找出的初步結果再經過第二步驟的處理修正影像灰階值。
為了判定實驗結果好壞，我們定義了一個可靠度的函數來決定此種內插方法的好壞。實驗結果顯示，我們所提出的方法優於傳統只使用一組資料的影像切片所做的線性內插，而在Z軸方向上也呈現出良好的內插影像結果。

In order to obtain better images of biological tissues, researches on microscopy imaging play a vital role in recent years. Confocal microscopy is widely used to capture images from biological specimens. To study the major mechanism of Drosophila brain activity, the brain science center (BRC) uses confocal microscopy to obtain images of Drosophila brain slices, hoping to construct a complete three-dimensional brain model. While the device plays a very important role, it has limits that restrict the quality of three-dimensional (3D) imaging. Due to limit of the specification, resolution is not equal in all coordinates. Neighboring pixels in xy-plane are closer than spacing between image slices. While reconstructing volume model using these image slices, inevitably resolution of z-axis will be lower than resolution of xy-plane.
In the thesis, we propose a super-resolution method. The interpolation method is used first to construct a relationship tree, and then to allocate close points for interpolation. Later, a rectification of the raw images will be applied to improve the image intensity qualities.
To judge the result of the interpolation, we define a reliability function to find out how reliable they are. The experimental result shows that it generate better reliability values than traditional linear interpolation and provides good z-axis information.

Chapter 1
1.1 Background
1.2 Motivation
1.3 Related Work
1.4 Thesis Organization
Chapter 2
2.1 Framework
2.2 Method Overview
Chapter 3
3.1 Introduction
3.2 Definition and Notation of k-d tree
3.3 Construction of k-d tree
3.4 Point-searching Method
3.5 Summary
Chapter 4
4.1 Introduction
4.2 Type of Rectification Function
4.3 Summary
Chapter 5
5.1 Reliability measurement
5.2 Reliability Results
5.3 Interpolation Results
5.4 Discussion and Analysis
Chapter 6
References

[1] Denis Segwogerere, Eric R. Weeks, “Confocal Microscopy,” Emory University, Atlanta, Georgia, U.S.A.
[2] George J. Grevera, Jayaram K. Upuda, “An Objective Comparison of 3-D Image Interpolation Methods,” IEEE Transactions on Medical Imaging, Vol. 17, No. 4, August 1998
[3] G. P. Penney, J. A. Schnabel, D. Rueckert, M. A. Viergever, W. J. Niessen, “Registration-Based Interpolation,” IEEE Transactions on Medical Imaging, Vol. 23, No. 47, July 2004
[4] Tong-Yee Lee, Chao-Hung Lin, “Feature-Guided Shape-Based Image Interpolation,” IEEE Transactions on Medical Imaging, Vol. 21, No. 12, December 2002
[5] Tong-Yee Lee, Wen-Hsiu Wang, “Morphology-Based Three-Dimensional Interpolation,” IEEE Transactions on Medical Imaging, Vol. 19, No. 7, July 2000
[6] Paul J. Besl, “A Method for Registration of 3-D Shapes,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 14, No. 2, February 1992
[7] Michael Greenspan, Mike Yurick, “Approximate K-D Tree Search for Efficient ICP,” Dept. of Electrical and Computer Engineering, School of Computing, Queen’s University, Kingston, Ontario, Canada
[8] Szymon Rusinkiewicz, Marc Levoy, “Efficient Variants of the ICP Algorithm,” In 3DIM01: 3D Imaging and Modelling,pages 145–152, 2001
[9] Jon Louis Bentley, “Multidimensional Binary Search Tree Used for Associated Searchng,” Stanford University, 1975
[10] Jerome H. Friedman, Jon Louis Bentley, Raphael Ari Finkel, “An Algorithm for Finding Best Matches in Logarithmic Expected Time,” ACM Transactions on Mathematical Software, Vol. 3, No. 3, pages 209-226, September 1977