JPEG Backward Compatible HDR Image Coding using Curve Fitting Prediction for DCT Coefficients : DCT 계수에 대한 커브피팅 예측을 이용한 JPEG 역방향 호환 가능한 HDR 이미지 코딩

다국어 초록 (Multilingual Abstract)

Currently, the digital imaging industry has begun the transition from low dynamic range (LDR) images using the traditional 8 bits per RGB color representation to high dynamic range (HDR) images, which use higher bit depths to represent each RGB color. The JPEG standard (ISO/IEC 10918-1) dominates the LDR image-coding market but offers no support for HDR images. To solve this problem, the JPEG committee (ISO/IEC JTC1 SC29 WG1) initiated a new HDR image-coding standardization, called JPEG XT (ISO/IEC 18477), which provides JPEG backward compatibility. In JPEG XT, HDR images are encoded into base- and residual-layer codestreams. An input HDR image is tone-mapped to an LDR image, which is compressed by the JPEG encoder, and a base-layer codestream is constructed to provide JPEG backward compatibility. The residual layer codestream allows a decoder to reconstruct an HDR version of the LDR codestream. To date, three profiles have been proposed for JPEG XT to encode HDR images which are represented using floating-point values. In this dissertation, a new novel image-coding scheme is proposed, which uses a block-based region adaptive prediction method, with a minor modification of the current specifications on JPEG XT Part 7. The proposed method predicts the HDR pixel value based on the relationship between the LDR and HDR blocks, whereas the existing JPEG XT profiles approximate the HDR values globally using a global inverse transformation of the LDR values. Experimental results using various sample images show the superiority of the proposed method in terms of objective evaluations and visual comparison.

목차 (Table of Contents)

• Abstract 1
• Preface 3
• Table of Contents 4
• List of Tables 7
• List of Figures 8
• I. Introduction and Overview 11
• 1. Introduction 11
• 2. High Dynamic Range Imaging Overview 14
• 2.1. High Dynamic Range Image Generation 16
• 2.1.1. Computer renderings 16
• 2.1.2. Fusing multi-exposed images 17
• 2.2. Encoding High Dynamic Rage Image 19
• 2.2.1. Lossless HDR Image Format 19
• 2.2.2. Lossy HDR Image Coding Standard 21
• 2.3. High Dynamic Range Image Display 22
• 2.3.1. Tone mapping 22
• 2.3.2. HDR Display 23
• 3. Dissertation Motivations and Objectives 24
• 4. Dissertation Contributions 26
• II. Evaluation of Backward Compatible HDR Image Coding Standard 28
• 1. JPEG XT: Backward Compatible HDR Image Coding Standard 28
• 2. Evaluation of Various Tone Mapping Operator for JPEG XT standard 33
• 2.1. Introduction 33
• 2.2. Related Considerations 34
• 2.2.1. Image Quality Assessment for Tone Mapped LDR image 34
• 2.2.2. No Reference Image Quality Assessment 35
• 2.3. Evaluations 37
• 2.3.1. Evaluation Software 37
• 2.3.2. Encoding Parameters 38
• 2.3.3. Datasets 39
• 2.3.4. Tone Mapping Operators 39
• 2.4. Experimental Results 40
• 2.4.1. Results of TMQI Measurement 41
• 2.4.2. Results of NR IQA Measurement 41
• 2.4.3. Results of PSNR Measurement 43
• 3. Performance Evaluation of the JPEG XT standard 46
• 3.1. Introduction 46
• 3.2. Experimental Setup 47
• 3.2.1. Tone Mapping Operator 47
• 3.2.2. Quality Metric 47
• 3.2.3. Dataset 48
• 3.2.4. Encoding Parameter 49
• 3.3. Experimental Results and Discussion 50
• 4. Chapter Conclusion 52
• III. Proposed HDR Image Coding Scheme 58
• 1. Problem Statement 58
• 2. Proposed Scheme 60
• 2.1. Basic Structure 60
• 2.2. HDR Prediction in DCT Domain 61
• 2.3. Suggestion for the JPEG XT standards 68
• 2.3.1. HDR Prediction in Spatial Domain 68
• 2.3.2. Decoding process of the proposed method 71
• 2.3.3. Modification of the Parametric Curve box 74
• 2.3.4. Implementation guide for the proposed scheme (Encoder) 77
• 2.3.5. Composition of the JPEG XT Segments for The Proposed Scheme 80
• 3. Experimental Results and Discussion 81
• 4. Chapter Conclusions 98
• IV. Dissertation Summary 100
• 1. Significance of the Research 100
• 2. Suggestions for Future Research 102
• References 103