Image processing and Machine Learning are used in conjunction to analyze and understand images. Where image processing is used to pre-process images using techniques such as filtering, segmentation, and feature extraction, Machine Learning algorithms are used to interpret the processed data through classification, clustering, and object detection. This book serves as a textbook for students and instructors of image processing, covering the theoretical foundations and practical applications of some of the most prevalent image processing methods and approaches. Divided into two volumes, this second installment explores the more advanced concepts and techniques in image processing, including morphological filters, color image processing, image matching, feature-based segmentation utilizing the mean shift algorithm, and the application of singular value decomposition for image compression. This second volume also incorporates several important Machine Learning techniques applied to image processing, building on the foundational knowledge introduced in Volume 1. Machine Learning (ML) is a branch of artificial intelligence (AI) that enables systems to learn from data and make informed predictions or decisions without the need for explicit programming. ML finds extensive applications in various domains. For instance, in automation, ML algorithms can automate tasks that would otherwise rely on human intervention, thereby reducing errors and enhancing overall efficiency. Predictive analytics is another area where ML plays a crucial role. By analyzing vast datasets, ML models can detect patterns and make predictions, facilitating applications such as stock market analysis, fraud detection, and customer behavior analysis. We have observed that students grasp the material more effectively when they have access to code that they can manipulate and experiment with. In line with this, our book utilizes MATLAB as the programming language for implementing the systems.
Author(s): Erik Cuevas, Alma Nayeli Rodríguez
Publisher: CRC Press
Year: 2023
Language: English
Pages: 239
Cover
Half Title
Title Page
Copyright Page
Table of Contents
Preface Volume II
1 Morphological Operations
1.1 Shrinkage and Growth of Structures
1.1.1 Neighborhood Types Between Pixels
1.2 Fundamental Morphological Operations
1.2.1 The Structure of Reference
1.2.2 Point Set
1.2.3 Dilation
1.2.4 Erosion
1.2.5 Properties of Dilatation and Erosion
1.2.6 Design of Morphological Filters
1.3 Edge Detection in Binary Images
1.4 Combination of Morphological Operations
1.4.1 Opening
1.4.2 Closing
1.4.3 Properties of the Open and Close Operations
1.4.4 The Hit-or-Miss Transformation
1.5 Morphological Filters for Grayscale Images
1.5.1 Reference Structure
1.5.2 Dilation and Erosion for Intensity Images
1.5.3 Open and Close Operations with Grayscale Images
1.5.4 Top-Hat and Bottom-Hat Transformation
1.6 MATLAB Functions for Morphological Operations
1.6.1 Strel Function
1.6.2 MATLAB Functions for Dilation and Erosion
1.6.3 MATLAB Functions Involving the Open and Close Operations
1.6.4 The Transformation of Success or Failure ('Hit-or-Miss')
1.6.5 The bwmorph Function
1.6.6 Labeling of Convex Components
Notes
References
2 Color Images
2.1 RGB Images
2.1.1 Composition of Color Images
2.1.2 Full-Color Images
2.1.3 Indexed Images
2.2 Histogram of an RGB Image
2.2.1 Histogram of RGB Images in MATLAB
2.3 Color Models and Color Space Conversions
2.3.1 Converting an RGB Image to Grayscale
2.3.2 RGB Images without Color
2.3.3 Reducing Saturation of a Color Image
2.3.4 HSV and HSL Color Model
2.3.5 Conversion From RGB to HSV
2.3.6 Conversion From HSV to RGB
2.3.7 Conversion From RGB to HLS
2.3.8 Conversion From HLS to RGB
2.3.9 Comparison of HSV and HSL Models
2.4 The YUV, YIQ, and YCbCr Color Models
2.4.1 The YUV Model
2.4.2 The YIQ Model
2.4.3 The YC[sub(b)]C[sub(r)] Model
2.5 Useful Color Models for Printing Images
2.5.1 Transformation From CMY to CMYK (Version 1)
2.5.2 Transformation From CMY to CMYK (Version 2)
2.5.3 Transformation From CMY to CMYK (Version 3)
2.6 Colorimetric Models
2.6.1 The CIEXYZ Color Space
2.6.2 The CIE Color Diagram
2.6.3 Lighting Standards
2.6.4 Chromatic Adaptation
2.6.5 The Gamut
2.7 Variants of the CIE Color Space
2.8 The CIE L*a*b* Model
2.8.1 Transformation CIEXYZ → L*a*b*
2.8.2 Transformation L*a*b* → CIEXYZ
2.8.3 Determination of Color Difference
2.9 The sRGB Model
2.10 MATLAB Functions for Color Image Processing
2.10.1 Functions for Handling RGB and Indexed Images
2.10.2 Functions for Color Space Conversion
2.11 Color Image Processing
2.12 Linear Color Transformations
2.12.1 Linear Color Transformation Using MATLAB
2.13 Spatial Processing in Color Images
2.13.1 Color Image Smoothing
2.13.2 Smoothing Color Images with MATLAB
2.13.3 Sharpness Enhancement in Color Images
2.13.4 Sharpening Color Images with MATLAB
2.14 Vector Processing of Color Images
2.14.1 Edge Detection in Color Images
2.14.2 Edge Detection in Color Images Using MATLAB
Note
References
3 Geometric Operations in Images
3.1 Coordinate Transformation
3.1.1 Simple Transformations
3.1.2 Homogeneous Coordinates
3.1.3 Affine Transformation (Triangle Transformation)
3.1.4 Projective Transformation
3.1.5 Bilinear Transformation
3.1.6 Other Nonlinear Geometric Transformations
3.2 Reassignment of Coordinates
3.2.1 Source-Destination Mapping
3.2.2 Destination-Source Mapping
3.3 Interpolation
3.3.1 Simple Interpolation Methods
3.3.2 Ideal Interpolation
3.3.3 Cubic Interpolation
3.4 Aliases
3.5 Functions for Geometric Transformation in MATLAB
3.5.1 Application Example
References
4 Comparison and Recognition of Images
4.1 Comparison in Grayscale Images
4.1.1 Distance between Patterns
4.1.2 Distance and Correlation
4.1.3 The Normalized Cross-Correlation
4.1.4 Correlation Coefficient
4.2 Pattern Recognition Using the Correlation Coefficient
4.2.1 Implementation of the Pattern Recognition System by the Correlation Coefficient
4.3 Comparison of Binary Images
4.3.1 The Transformation of Distance
4.3.2 Chamfer Algorithm
4.4 Chamfer Index Relationship
4.4.1 Implementation of the Chamfer Relation Index
References
5 Mean-Shift Algorithm for Segmentation
5.1 Introduction
5.2 Kernel Density Estimation (KDE) and the Mean-Shift Method
5.2.1 Concentration Map Generation
5.3 Density Attractors Points
5.4 Segmentation with Camshift
5.4.1 Feature Definition
5.4.2 Operative Data Set
5.4.3 Operation of the MS Algorithm
5.4.4 Inclusion of the Inactive Elements
5.4.5 Merging of Not Representative Groups
5.4.6 Computational Process
5.5 Results of the Segmentation Process
5.5.1 Experimental Setup
5.5.2 Performance Criterion
5.5.3 Comparison Results
References
6 Singular Value Decomposition in Image Processing
6.1 Introduction
6.2 Computing the SVD Elements
6.3 Approximation of the Data Set
6.4 SVD for Image Compression
6.5 Principal Component Analysis
6.6 Principal Components through Covariance
6.7 Principal Components through Correlation
References
Index
