De-interlacing Algorithm Using Classification of Image Feature

在過去的日子裡，交錯式的訊號廣泛的使用。而隨著高科技時代的來臨，採用交錯式或漸進式影像訊號的問題也被討論相當多年。相較於傳統的類比電視，高解析度及漸進式掃描是一種趨勢。因此，把電視訊號轉換成漸進式訊號是很重要的一門技術。而這個處理的過程就稱之為解交錯。總的來說，當我們要把交錯式的訊號內容呈現在漸進式顯示裝置上時，解交錯是不可缺少的一個步驟。 整篇論文分為四個章節。在第一章裡，我們先對解交錯做簡單的介紹。在第二章會描述一些以前所提出的演算法，並且把這些演算法分為四大類。而我們所提出來的方法則會在第三章做詳細的說明，內容包括動態偵測以及以影像特徵為基底的內插方法，同時也會提供實驗的結果。第四章是全文的總結，在此也提出未來可以改進的方向。The interlaced signals were widely used in past. With the coming of high technology age, the question to adopt interlacing or progressive video signal is discussed for many years. Comparing the traditional analog TV display, large size/high resolution and progressive scanning is the trend. Hence, it is an important technique to convert ordinary TV sequences into progressive sequences. The process is called as de-interlacing. To sum up, de-interlacing is an indispensable step while presenting interlaced content on progressive display devices. This dissertation includes four chapters. In chapter 1, we give a brief introduction of de-interlacing. Some previous works are described in chapter 2, and we divide these into four categories. A detailed description of our proposed method is presented in chapter 3, including motion detector and feature based interpolation. Experiment results will also be provided. In chapter 4, we will conclude our researches and the future works will also be discussed here.Chapter 1 Introduction ………………………………………………….. 1 1.1 Organization ………………………………………………………………. 1 1.2 Background and motivation …………………………………………….… 2 1.2.1 What and Why “interlacing” ? ………………………………….… 3 1.2.2 What and Why “de-interlacing” ? ……………………………….... 5 1.3 Summary ………………………………………………………………….. 7 Chapter 2 Previous de-interlacing algorithms ………………………….. 9 2.1 Spatial de-interlacing algorithm ……………………………………….… 11 2.1.1 Double line interpolation …………………………………….... 11 2.1.2 Linear filtering interpolation ………………………………..… 12 2.1.3 Edge dependent interpolation ………….……...………………. 13 2.2 Spatial-Temporal de-interlacing algorithm ……………………………… 18 2.2.1 Inter-field merge interpolation ………………………………... 18 2.2.2 Inter-field average interpolation ………………………………. 20 2.2.3 Vertical-Temporal median filtering interpolation ……………... 21 2.3 Motion adaptive de-interlacing algorithm ……………..……………….... 22 2.4 Motion compensation de-interlacing …………………………………..… 23 2.5 Summary ………………………………………………...………………. 25 Chapter 3 The proposed algorithm and simulation results ……………. 27 3.1 Motion detector ………………………………………………………….. 28 3.2 Feature-based average interpolation ………………..…...……………….. 31 3.3 Simulation results ………………………………………………………... 42 3.4 Summary ………………………………………………………………… 52 Chapter 4 Conclusion …………………………………………………. 55 Reference ……………………………………………………………… 5