基於線性融合之影像對比強化方法

Image Contrast Enhancement Based on Linear Fusion

陳柏璋(Bo-chang Chen)

指導教授 ： 謝政勳

朝陽科技大學/資訊學院/資訊工程系/碩士(2011年)

摘要

傳統直方圖均化(conventional histogram equalization, CHE)是一個常見的影像強化方法，它能夠有效強化影像的對比與細節，但是容易出現過度強化的問題，雖然如此，原始影像在經過CHE後，一些在原始影像中不明顯的細節會出現在CHE影像中，這代表原始影像與CHE影像有互補的現象，我們稱此現象為影像細節互補特性(detail complementary property, DCP)。本論文將利用DCP特性將原始影像與CHE圖以線性內插融合方式提升影像視覺品質。
本篇論文中，基於線性融合我們提出二種影像對比強化方法。一種是採用固定融合參數的方法，稱為DACE/LIF (detail aware contrast enhancement with linear image fusion)，使用固定參數於影像融合過程。另一種是採用適應性融合參數的方法，稱為APLIF (adaptive pixel-based linear fusion)，其中使用標準差來取得適應性融合參數。為了驗證本論文所提方法的優異性，我們與不同的影像強化方法進行比較。結果顯示本論文的方法在對比強化上均優於比較方法。

關鍵字

直方圖均化 ； 影像融合 ； 影像強化 ； 對比強化

並列摘要

Conventional histogram equalization (CHE) is a common image enhancement method, which can effectively enhance image contrast and details. However, one of fundamental problems in CHE is over enhancement. Even though, it is observed that the equalized image by CHE reveals details which are not clear in the original image while loses details in some other parts of original image. Interesting enough, the details missing in the equalized image can be found in the original image and the details not clear in the original image can be obtained from the equalized image. We call this property as detail complementary property (DCP). By the DCP, two approaches to image contrast enhancement based on linear fusion are presented in this thesis. They are DACE/LIF (detail aware contrast enhancement with linear image fusion) and APLIF (adaptive pixel-based linear image fusion).
With an image and its CHE equalized image, a linear image fusion is used to obtain a fused image whose contrast and visual quality are enhanced. The linear image fusion is chosen in this thesis because the two images are of the DCP. When the fusion parameter is fixed, it is the DACE/LIF approach. When the fusion parameter is adaptively calculated, it is the APLIF approach.
Both DACE/LIF and APLIF approaches are verified by several examples. The results indicate that the two approaches are able to achieve good performance of contrast enhancement. Besides, the two approaches are compared with three enhancement approaches. The results are for the proposed approaches since better contrast and visual quality are found in enhanced images.

並列關鍵字

histogram equalization ； contrast enhancement ； image fusion ； image enhancement

參考文獻

[1] Yeong-Taeg Kim, ”Contrast enhancement using brightness preserving bi-histogram equalization” IEEE Transactions on Consumer Electronics, Vol.43, No.1, pp. 1– 8, November 1997.

[2] Qing Wang, Rabab Ward, “Fast image/video contrast enhancement based on weighted thresholded histogram equalization,” IEEE Transactions on Consumer Electronics, Vol.53, No.2, pp.757–764, May 2007.

[3] Chen Hee Ooi, Pik Kong, Nicholas Sia, Ibrahim, Haidi, “Bi-histogram equalization with a plateau limit for digital image enhancement,” IEEE Transactions on Consumer Electronics, Vol. 55, No. 4, pp. 2072–2080, November 2009.

[4] How-Lung Eng and Kai-Kuang Ma, ”A dynamic histogram equalization for image contrast enhancement,” IEEE Transactions on Consumer Electronics, Vol. 53, No.2, pp. 593–600, May 2007.

[5] Beilei Xu, Yiqi Zhuang, “Object-based multilevel contrast stretching method for image enhancement,” IEEE Transactions on Consumer Electronics, Vol. 56, No. 3, pp. 1746–1754, August 2010.