FocusClear?處理過後之透明生物顯微鏡影像的三維反疊積

自從顯微鏡發明以來，一直是醫學界和生物科技界重要的設備之一。對醫學界而言，能夠利用顯微鏡觀察病人的生理組織，有助於醫生更準確的判斷病情，減少錯誤判斷率。對生物科技而言，顯微鏡使能見微知著，讓科學家了解微觀世界的脈動，才能更詳細了解生物的運作，對現代科技的發展幫助很大。顯微鏡的種類很多，解析度也各自不同，但最終目的不外乎希望能夠看到更清楚更細節的影像，當然價錢也是其中考量。本研究的目的是利用影像演算法以期增加顯微鏡的三維影像解析度。更進一步希望研發出低價卻高解析度的顯微鏡。影響解析度的原因有很多，擷取影像時的背景光、感光器本身的雜訊、相機的畫素、甚至樣本自身來自不同焦距平面的模糊影像都會決定最後影像的品質。這份論文主要是針對共軛焦螢光顯微鏡，擷取生物組織的三維變焦影像(一系列二維影像堆疊)，經由影像處理的三維反疊積(Deconvolution)後產生高解析度的三維生物影像。在這整個演算法中，本論文著重於生物細胞組織經由光學顯微鏡擷取的影像，由於鏡頭的不完美而造成的模糊變型現象，利用盲反疊積(Blind Deconvolution)估計出的點擴散方程式(Point Spread Function)實行於整體三維影像中完成三維影像重建，減輕原本失焦所造成的模糊和偏光所產生的形變問題。

關鍵字

反疊積；顯微鏡；三維

並列摘要

Since the invention of microscope, it has been one of the most important equipments in medical and biological technology communities. For medical applications, doctors investigate patients’tissues through microscopes, which help them better diagnose disease and decrease possibility of errors. For biologists, microscopes help them recognize things through micro world, understand how living creature works in detail. In summary, microscopy has great advantage to nowaday technology development. There are many kinds of microscopes, each with different resolutions, but eventually they are all in essence developed to observe images more clearly with more details. Price, of course, is one of the concerns. This research is intended to increase the resolution of 3D microscopy images by image processing algorithms. The factors affecting the captured resolution can be complex, including the backlight of the scene, noises of the sensor, pixel rate of the camera, and even the sample itself affects the final resolution due to defocus. This thesis is mainly focused on confocal fluorescence microscopy, capturing defocused 3D images (a series of 2D image stack), and then further image processing on the 3D image with 3D deconvolutionto to produce the 3D super-resolution sample. In the whole algorithm, this study focuses on processing the images of biological tissue from light microscopes. Since the imperfection of lenses results in blurriness and distortion of the original sample, Blind Deconvolution will be employed to estimate Point Spread Function on the raw 3D data for subsequent 3D reconstruction to alleviate these problems.

並列關鍵字

Deconvolution ； Microscopy ； 3D

參考文獻

[2]T. F. Brooks and W. M. Humphreys, “A deconvolution approach for the mapping of acoustic sources (DAMAS) determined from phased microphone arrays,” Journal of Sound and Vibration, Volume 294, Issues 4-5, pp. 856-879, 2006.

[3]T.J. Holmes, “Acceleration of maximum-likelihood image restoration for fluorescence microscopy and other noncoherent imagery,” Journal of Optical Society of America A, Volume 8, Issue 6, pp. 893-907, 1991.

[4]T.J. Holmes and N. J. O’connor, “Blind deconvolution of 3D transmitted light brightfield micrographs,” Journal of Microscopy, Volume 200, Issue 2, pp. 114-127, 2000.

[5]T.J. Holmes, “Blind deconvolution of quantum-limited incoherent imagery: maximum-likelihood approach,” Journal of Optical Society of America A, Volume 9, Issue 7, pp. 1052-1061, 1992.

[7]S. Kimura and C. Munakata, “Dependence of 3-D optical transfer functions on the pinhole radius in a fluorescent confocal optical microscope,” Applied Optics, Volume 29, Issue 20, pp. 3007-3011, 1990.

國際替代計量

FocusClear?處理過後之透明生物顯微鏡影像的三維反疊積

全文下載

主題瀏覽