應用深度資訊於影像增強之3D算圖研究

A study of Image Enhancement for 3D Rendering Through Depth Information

蔡奉迪(Feng-Ti Tsai)

指導教授 ： 黃銘智

國立臺北科技大學/設計學院/創新設計研究所/碩士(2008年)

https://doi.org/10.6841/NTUT.2008.00262

摘要

在由3D軟體所產生的圖像中，照明是一個很重要的因素。然而應用於實務上，3D物件往往非常複雜，照明的參數設定不易，可能因為照明不足而太暗，造成影像不清；也可能照亮了某處，卻使得其他部份過亮。同時還需要許多專業知識才能達成，例如需要許多燈光及參數設定等等。所以以現有的技術產生影像時，在品質與運算時間上可能有不容易兼顧的問題。

目前人們對於圖像的視覺品質要求越來越高，而一方面又希望軟體在使用上能夠越來越容易且快速。因此本研究提出一套簡易且有效率的方式，進行影像增強，使場景暗處之細節更容易判讀。本文以深度資訊(Depth Information)為基礎，採用濾色(Screen)之運算方式，利用深度資訊中含有物件輪廓細節，以及不受照明影響之特性，對圖像色彩中之明度值進行不等量的提升，增加圖像中的顏色層次，豐富圖像所提供之視覺資訊。不需要大量光源照明，便可解決物件因照明不足，造成輪廓與細節不清的問題。

文中利用三個照明不足之案例，比較增強前後之明度變化曲線。証明以所提之方法確實可保留亮部細節，並且呈現暗部的細節。同時場景中，不需要大量的光源設定。

關鍵字

影像增強 ； 深度資訊 ； 3D算圖

並列摘要

Illumination is one of the most important element in 3D CAD systems, but complex scenes are not easy to illuminate. Usually we make some place in the scene illuminated, but other places overexposed at the same time. Users also have to understand much professional knowledge to use 3D software – like setting parameters or lights, so it is hard to keep both quality and render time by existed CG technology.
Nowadays, standard CG quality is getting higher and higher, but people still want using software easily, so we present a simple and efficient method to enhance images by depth information. Depth information will not effected by illumination, and due to “Screen” calculation, it could raise the brightness value of every pixel unequally to enhance the gradation of images. So this method could make somewhere too dark on the image become visible, without using lots of lights.
Then this paper compared the brightness value curves of original and enhanced image, to test our method. The result is our method could enrich the information of dark parts, and keep the information of bright parts at the same time. Besides, our method does not need to create lots of lights in the scene, so this method both reduced the procedure when using software and enriched image information.

並列關鍵字

Image Enhancement ； Depth Information ； 3D Rendering

參考文獻

[5] Carlson, Wayne. 2003. A Critical History of Computer Graphics and Animation. The Ohio State University. http://design.osu.edu/carlson/history/lessons.html.

[8] Hummel, R. A. 1975. Histogram modification techniques. Computer Graphics and Image Processing. 1975, 3.

[9] Landis, Hayden. 2002. Production-Ready Global Illumination. SIGGRAPH 2002.

[10] Luft, Thomas, Colditz, Carsten and Deussen, Oliver. 2006. Image Enhancement by Unsharp Masking the Depth Buffer. SIGGRAPH 2006.

[17] Smith, Alvy Ray. 1978. Color Gamut Transform Pairs. Computer Graphics. Vol. 12, 3.