影像細節疊加對比演算法研究

Abstract

本論文針對影像亮度部份提出一套可提高對比度同時改善影像細節的方法，並以硬體電路加以實現完成。本論文有三大特點第一是採用亮度及色彩分離較佳的CIEL*a*b*色彩空間轉換，第二是影像處理方式由細節調整及對比調整疊加完成，第三是本論文演算法可用於動態影像處理。 本論文採用CIEL*a*b*色彩空間是要得到較符合於人眼的影像，且較容易以線性方式分析、處理影像。影像細節疊加對比架構採用互補式設計，此設計可同時保留影像細節及加強影像對比。本論文演算法以動態分析影像及參數調整，可加強所有動態輸入影像。 根據理論基礎及硬體電路驗證，證實本論文所提出的方法確有其可行性與利用價值，也期待能提供影像研究者不同的想法。

In this thesis, an advanced method which can improve detail and contrast of image in luminance is presented and associated hardware experimentation is developed. There are three major features in this thesis ,the first feature is CIEL*a*b* color space transformation which has better separation of luminance and chroma, the second is image processing method which is implemented by detail and contrast cascade, and the third is dynamic image processing. This thesis uses CIEL*a*b* color space which has proper perceptual image and it also can be easily analyzed and processed in linear format. Image detail and contrast cascade architecture has complementary design which could preserve detail and enhance contrast of image simultaneously. The presented approach of dynamic image analysis and parameter adjustment could help enhance the dynamic input images. From the theoretical basis and associated hardware implementation, the presented approach in this thesis has shown feasible and usable. We also expect this content could stimulate different thinking in image research.