互動式與半自動化參數輪廓擷取; Interactive and Semi-automatic Parametric Contour Extraction

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/8850

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/8850

Title:	互動式與半自動化參數輪廓擷取;Interactive and Semi-automatic Parametric Contour Extraction
Authors:	劉堂輝;Tang-Hui Liu
Contributors:	資訊工程研究所
Keywords:	主動輪廓模式;輪廓擷取;active contour model;contour extraction
Date:	2004-06-25
Issue Date:	2009-09-22 11:36:16 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	影像編輯是一個專業性的工作，具備影像相關知識的人才能做好這個工作。一般不具備影像專業知識的人僅能依靠特殊軟體才可能編輯好影像，就像傻瓜相機相對於不具備攝影知識的人一樣。我們的研究目的即是提供一套適合不具備影像專業知識的人使用的影像編輯軟體。本論文的研究內容是這個研究系列的第一步驟。本系統中的基本功能有：亮度、對比、及色彩飽和度的調整。我們將實作出幾種方法比較，找出最符合人性操作的方式。而本論文最主要研究在於互動式與半自動化參數輪廓的擷取。在人性化編輯系統中選取工具提供兩種選擇，一是互動式選取，我們利用Catmull-Rom spline 實作出一選取工具，使用者可以在想要選取的物體外圍點下適當的點，然後系統會在這些點中內插適當的點而形成一封閉曲線，如果不準確，我們可以動態改變點的位置或額外加點內插以提高輪廓的準確性。二是利用主動輪廓模式 (active contour model) 做半自動的輪廓選取。主動輪廓模式是利用能量公式，將圈出來的大概輪廓自動逼近到準確的位置及形狀。由於主動輪廓模式的複雜度相當高，本研究的重點在於改進其執行速度與效果。 Image edition is a professional work; only people who have the related knowledge of images can do it well and the edited results are dependent on the levels of their related knowledge. Anyone can edit images with image edit software; however, the edited results heavily depend on his/her related knowledge. Anyone who has no photography knowledge can take well pictures using an automatic camera; so we think that there is an image edit software making people who have no image knowledge to edit images well. In our study, we try to provide an image editor such that people who have no image knowledge can edit images well by using the editor. The work of this thesis is just the first step of this studying. This work has two major topics: one is the basic image editing operators and the other is the contour extraction. In the basic editing operations, several different-concept tuning operators on brightness, contrast, and saturation are proposed and compared. Contour extraction is our main task. Two contour extraction algorithms are developed: one is the interactive contour extraction and the other is the semi-automatic parametric contour extraction. Catmull-Rom splines are used for interactive parametric contour extraction. The control points for the splines can be added and moved to approach the real contour shape. Semi-automatic parametric contour extraction is based on the active contour models. An adaptive active contour model improved from the fast greedy algorithm is proposed. Experiments show that the proposed method can extract object boundary contour more accurate and efficient.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing

Loading...