此篇論文中,我們提出一個適用於TFT-LCD上之影像改善方式AICE(Adaptive Image Contrast Enhancement),使用者可簡單地選擇他們喜愛之改善程度,進而達到對比及輝度調整。 在我們提出可適性之對比增強方式,有幾個主要的重點。第一,TFT-LCD內含一個時序控制器接收數位RGB之輸入資料,且輸入之RGB數位資料同時連接到AICE機制,經過histogram後判斷輸入之影像分佈,獲得PDF及CDF;利用CDF之近似值得到轉移函數。第二,我們利用得到的轉移函數與階調參考電壓間的階調電壓mapping,產生programming signal控制multi-channel之DAC,進而將輸出給Source Driver之類比電壓作調整控制,達到影像改善之目的。在這篇論文中,我們提出以下的貢獻: 1. 我們可以獲得影像改善:我們利用提出的AICE方式,分析統計輸入影像資料之灰階分佈,可以獲得輸入與輸出影像之轉移函數,並由事先設定之三個主要參數α、β及γ作為預防影像過度增強之限制條件。 2. 我們可以保持時序控制器之數位資料:在舊有提出的內插方式中,數位資料畫素增強將會隨著解析度越高而產生不當的過度增強,使畫面失真。我們的AICE架構是將輸入影像信號另外作處裡後給Programming Multi-channel DAC作類比電壓調整,故原本時序控制器處理之影像資料不會因此改變受到影響。 最後,我們可以知道,我們提出的AICE(Adaptive Image Contrast Enhancement)可改善影像對比之品質,並在未來研究領域上,可作為動態畫面改善之基礎方法。
In this thesis, we propose a scheme of contrast enhancement to improve TFT-LCD image quality. Applying this technique in the TFT-LCD module, users could easily select the enhancement levels, and then the picture contrast and brightness could be adjusted adaptively. In our proposal, the new contrast enhancement has several main focuses. At first, the TFT-LCD display device contains a timing controller (TCON) ASIC that receives input digital image RGB (Red/Green/Blue) data from a graphics source. Our Adaptive Image Contrast Enhancement (AICE) arithmetic center which applying to the basis of Image Histogram. The image histogram included restricting factors also receives input digital image RGB data with the TCON synchronously. Then we also propose the desired input-output transfer function with Quantify Approximation of CDF, it maps the gamma reference voltage curve by Gamma Voltage Mapping schema. Finally, we propose the programming multi-channel DAC to control the gamma voltages of TFT-LCD due to the fixed R-String resistances not suit all input image. In this thesis, the following contributions are gained: 1. Adaptive Image: We adapt the method of Adaptive Image Contrast Enhancement (AICE) included restricting factors, to avoid the brightness excessive changing after histogram equalization. 2. Image Data Preserving: By our proposed image contrast enhancement method, we preserve the original image data in timing control process. To avoid existing interpolation algorithms that people have been proposed for digital image enlargement on all pixels in the size of FPD panels increase and need higher resolution. Finally, we know that our proposed method can improve the contrast enhancement to get higher image quality and also propose a future research in moving picture.