High Dynamic Range Video

Concepts, Technologies and Applications

Alan Chalmers, Patrizio Campisi, Peter Shirley, Igor García Olaizola

• Publisher: Academic Press
• Year: 2016
• Print ISBN: 9780128094778
• eBook ISBN: 9780081010389
• Topic: Technology & Engineering
• Subtopic: Digital Media
• Index: Technology & Engineering

High Dynamic Range Video: Concepts, Technologies and Applications gives an introduction to a full range of topics within the end-to-end HDR video pipeline, covering the issues around capturing HDR and stereo HDR video, such as ghosting and use of legacy LDR systems, how HDR video can be manipulated, including real-time mixing, the very latest designs for HDR displays, HDR video on mobile devices, and the applications of HDR video. With this book, the reader will gain an overview of the current state-of-the art of HDR video, learn the potential of HDR video to provide a step change to a wide range of imaging applications, and attain the knowledge needed to introduce HDR video in their own applications. - Written by experts who have been actively researching High Dynamic Range Video - Covers a full range of topics within the end-to-end HDR video pipeline - Provides applications that demonstrate how HDR video can be applied

Part 1

Capture

Chapter 1

Ghosting in HDR Video

A.O. Akyüz*; O.T. Tursun^†; J. Hasić-Telalović^‡; K. Karađuzović-Hadžiabdić^{‡ *} Middle East Technical University, Ankara, Turkey
^† Max Planck Institute for Informatics, Saarland, Germany
^‡ International University of Sarajevo, Sarajevo, Bosnia and Herzegovina

Abstract

HDR video capture is a very important problem, the solution of which can facilitate the transition to an end-to-end HDR video pipeline. This problem is currently being tackled using various approaches. On the hardware front, there exist designs with a judicious combination of multiple off-the-shelf components such as lenses, mirrors, beam-splitters, and sensors, as well as designs that involve high dynamic range sensors. On the software front, many HDR video deghosting algorithms exist that enable creation of HDR videos from differently exposed successive frames. These algorithms need to deal with motion within the captured scene, as well as the motion of the capture device to produce artifact-free HDR videos. This chapter provides an overview of the most notable algorithms in both approaches, with an emphasis on HDR video deghosting algorithms. For further reading, we refer the interested readers to other excellent resources on this topic.

Keywords

High dynamic range; Motion compensation; Deghosting; Global alignment; Local alignment

Contents

1. Introduction

2. Image Acquisition Model

3. HDR Image Deghosting Methods

4. HDR Video Capture Methods

4.1 Single Aperture Multiple Sensor Solutions

4.2 Multiple Camera Solutions

4.3 HDR Sensor Solutions

5. Global Video Deghosting

5.1 View Dependent Enhancement of the Dynamic Range of Video

5.2 Histogram-Based Image Registration for Real-Time HDR Videos

5.3 A Real-Time System for Capturing HDR Videos

6. Local Video Deghosting

6.1 High Dynamic Range Video

6.2 Video Matching

6.3 HDR Video Through Fusion of Exposure Controlled Frames

6.4 High Dynamic Range Video With Ghost Removal

6.5 Spatially Adaptive Filtering for Registration Artifact Removal in HDR Video

6.6 Filter-Based Deghosting for Exposure Fusion Video

6.7 Toward Mobile HDR Video

6.8 Patch-Based High Dynamic Range Video

6.9 A MAP Estimation Framework for HDR Video Synthesis

6.10 Motion Aware Exposure Bracketing for HDR Video References

1 Introduction

High-quality HDR videos can be captured in the following two ways: (1) by using dedicated video capture hardware that has improved dynamic range and (2) by using standard hardware to capture a set of frames with alternating exposures and then combining these frames for extending the dynamic range. We will call these two groups of techniques simultaneous and sequential HDR video capture methods.

Simultaneous HDR video capture typically entails the use of sensor elements with improved dynamic range. This method is usually employed by commercial products. However, it can also be performed by spatially varying the exposure of each sensor element, spatially varying the transmittance of neutral density filters placed in front of the sensor, or using a beam-splitting system to redirect the incoming light into multiple sensors, each set to a different exposure value. In this latter case, it becomes necessary to merge the output of the individual sensors to obtain an HDR frame.

Sequential HDR video capture, on the other hand, involves capturing of frames with alternating low and high exposures. This variation can be achieved by changing the exposure time, aperture size, or the sensitivity (ISO) of the camera between frames. In these techniques, each frame of the video is captured in low dynamic range (LDR), but has the potential to be converted to HDR by using the information present in the adjacent frames. These sequential techniques have recently become popular thanks to programmable digital cameras or third par...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Contributors
• Editors' Biography
• Acknowledgments
• Overview of HDR Video
• Part 1: Capture
• Part 2: Manipulate
• Part 3: Display
• Part 4: Applications and Uptake
• Index