Sea Ice Image Processing with MATLAB®

Qin Zhang, Roger Skjetne

• Publisher: CRC Press
• Year: 2018
• eBook ISBN: 9781351069182
• Topic: Technology & Engineering
• Subtopic: Civil Engineering
• Index: Technology & Engineering

Sea Ice Image Processing with MATLAB addresses the topic of image processing for the extraction of key sea ice characteristics from digital photography, which is of great relevance for Artic remote sensing and marine operations. This valuable guide provides tools for quantifying the ice environment that needs to be identified and reproduced for such testing. This includes fit-for-purpose studies of existing vessels, new-build conceptual design and detailed engineering design studies for new developments, and studies of demanding marine operations involving multiple vessels and operational scenarios in sea ice. A major contribution of this work is the development of automated computer algorithms for efficient image analysis. These are used to process individual sea-ice images and video streams of images to extract parameters such as ice floe size distribution, and ice types. Readers are supplied with Matlab source codes of the algorithms for the image processing methods discussed in the book made available as online material.

Features

Presents the first systematic work using image processing techniques to identify ice floe size distribution from aerial images

Helps identify individual ice floe and obtain floe size distributions for Arctic offshore operations and transportation

Explains specific algorithms that can be combined to solve various problems during polar sea ice investigations

Includes MATLAB® codes useful not only for academics, but for ice engineers and scientists to develop tools applicable in different areas such as sustainable arctic marine and coastal technology research

Provides image processing techniques applicable to other fields like biomedicine, material science, etc

Contents

Foreword

Preface

Acknowledgments

Contributors

List of Figures

List of Tables

Chapter 1 Introduction

1.1 Research background

1.2 Sea ice parameters

1.2.1 Ice concentration

1.2.2 Ice types

1.2.3 Ice floe size and floe size distribution

1.3 Applications of digital image processing techniques for ice parameter identification

1.3.1 Ice concentration calculation

1.3.2 Sea ice type classification

1.3.3 Ice floe identification

1.4 Book scope and structure

Chapter 2 Digital Image Processing Preliminaries

2.1 Image types

2.1.1 Grayscale image

2.1.2 Color image

2.1.2.1 The RGB color space

2.1.2.2 The CMY and CMYK color spaces

2.1.2.3 The HSI color space

2.1.3 Indexed image

2.2 Image histogram

2.3 Basic relationships between pixels

2.3.1 Pixel neighborhoods

2.3.2 Adjacency

2.3.3 Path

2.3.4 Connectivity

2.3.5 Region and boundary

2.4 Distance transform

2.4.1 Euclidean distance

2.4.2 City-block distance

2.4.3 Chessboard distance

2.4.4 Performance of the distance metrics

2.5 Convolution

2.6 Set and logical operations

2.6.1 Set operations on binary images

2.6.2 Set operations on grayscale images

2.6.3 Logical operations

2.7 Chain code

2.8 Image interpolation

2.8.1 Nearest neighbor interpolation

2.8.2 Bilinear interpolation

2.8.3 Bicubic interpolation

Chapter 3 Ice Pixel Detection

3.1 Thresholding

3.1.1 Global thresholding

3.1.1.1 Otsu thresholding

3.1.2 Local thresholding

3.1.3 Multithresholding

3.2 Clustering

3.2.1 Clustering types

3.2.2 K -means clustering

3.3 Experiment results and discussion

Chapter 4 Ice Edge Detection

4.1 Derivative edge detection

4.1.1 Gradient operator

4.1.2 Laplacian

4.2 Morphological edge detection

4.2.1 Erosion and dilation

4.2.1.1 Binary erosion and dilation

4.2.1.2 Grayscale erosion and dilation

4.2.2 Morphological closing and opening

4.2.2.1 Morphological closing

4.2.2.2 Morphological opening

4.2.3 Morphological reconstruction

4.2.3.1 Binary morphological reconstruction

4.2.3.2 Grayscale morphological reconstruction

4.2.4 Morphological gradient

4.3 Experimental results and discussion

Chapter 5 WatershedBased Ice Floe Segmentation

5.1 Watershed segmentation

5.1.1 Watershed segmentation using gradients

5.1.2 Watershed segmentation using the distance transform

5.1.3 Marker-controlled watershed segmentation

5.2 Combination of the watershed and neighboring-region merging algorithms

5.2.1 Concave detection by chain code

5.2.1.1 Boundary tracing

5.2.1.2 Differential chain code

5.3 Experimental results and discussion

Chapter 6 GVF SnakeBased Ice Floe Boundary Identification and Ice Image Segmentation

6.1 Traditional parametric snake model

6.1.1 The energy functionals

6.1.1.1 Internal energy

6.1.1.2 External energy

6.1.2 Implementation

6.1.3 Limitations

6.2 Gradient vector flow (GVF) snake

6.3 Contours initialization for applying the GVF snake algorithm in ice floe boundary identification

6.3.1 The location of initial contour

6.3.2 The shape and size of the initial contour

6.3.3 Automatic contour initialization based on the distance transform

6.4 Ice image segmentation

6.5 Discussion

6.5.1 Stopping criterion for the snake

6.5.2 GVF capture range

6.5.3 Border effects

Chapter 7 Sea Ice Type Identification

7.1 Ice shape enhancement

7.1.1 Morphological cleaning

7.1.2 Connected component extraction and labeling

7.1.3 Hole filling

7.1.3.1 Iterative dilation-based hole filling algorithm

7.1.3.2 Morphological reconstru...

Table of contents

• Cover
• Halftitle
• Title Page
• Copyright Page
• Table of Contents