- 學位論文
植基於高動態範圍結構之二維凝膠電泳影像相關研究
Researches of 2DEG Images based on High Dynamic Range
摘要
二維凝膠電泳影像(Two-Dimensional Electrophoresis Gel Image, 2DEG Image)是一種用來偵測與比對蛋白質點的重要生物影像。但二維凝膠電泳影像容易受限於截取設備的低動態範圍影響，導致二維凝膠電泳工具無法有效地正確偵測出微量蛋白質點。本論文針對二維凝膠電泳影像做出一系列的相關研究，希望能協助生物學家進行蛋白質研究，有效整合工作，提升研究效率。 二維凝膠電泳影像的研究成果可分成三大部分，內容如下。首先是以程式批次設定，取代人工手動，完成二維凝膠電泳影像批次截取的工作，達到省時與避免人為因素影響影像品質的優點。接著利用所得的影像，以類似高動態範圍影像 (High Dynamic Range Image)的觀念，設計出HDB (High Dynamic Bit)，以提升影像的位元數，完整保留影像在不同明亮度的特徵；經由實驗證明，可以幫助二維凝膠電泳工具找出原本偵測不到的微量蛋白質。最後為了有效保存HDB這類高位元影像，LSHDBC(Lossless High Dynamic Bit Compression)以JPEG-LS為基礎，結合基因演算法的觀念，隨著影像位元數的變化，調整相關參數，與其它一般檔案壓縮程式相比之下，具有良好的壓縮效果。
並列摘要
Two-Dimensional Electrophoresis Gel (2DEG) image is a popular method used in protein studies. However, 2DEG images could be affected by the process of image interception, lead to an inaccurate detection results. In order to solve the problem, this paper is aimed at 2DEG images, proposed a serious of research methods, assist biologists to engage in 2DEG image of the relevant researches, effective integration work, enhance the efficiency of researches. In this paper, the research results of 2DEG images can be divided to three parts. First, we designed a program to comply with automatic batch scanning, decrease the time and manpower of getting digital images；next is improving the accuracy of protein detection, we proposed HDB (High Dynamic Bit), through raising the bits number of image, preserve the characteristics of the images in the different brightness. In our experimental result, HDB can help the tool of 2DEG to detect more protein. Finally, in order to overcome the problem of saving HDB which is an high dynamic bit image, LSHDBC(Lossless High Dynamic Bit Compression) based on JPEG-LS and combine with genetic algorithm, according to bit number of image, adjust the relevant parameters of JPEG-LS, the file size of LSHDBC is smaller than JPEG-LS and other general file compression program.