Image Processing Lecture PDF

Image processing lecture Dr.Hazim G. Daway CHAPTER 1 INTRODUCTION AND OVERVIEW WHAT WILL WE LEARN? What is image processing? What are the main applications of image processing? What is an image? What is a digital image? What are...

Image processing lecture Dr.Hazim G. Daway CHAPTER 1 INTRODUCTION AND OVERVIEW WHAT WILL WE LEARN? What is image processing? What are the main applications of image processing? What is an image? What is a digital image? What are the goals of image processing algorithms? What are the most common image processing operations? Which hardware and software components are typically needed to build an image processing system? What is a machine vision system (MVS) and what are its main components? Why is it so hard to emulate the performance of the human visual system (HVS) using cameras and computers? 1.1 MOTIVATION Humans have historically relied on their vision for tasks ranging from basic instinctive survival skills to detailed and elaborate analysis of works of art. Our ability to guide our actions and engage our cognitive abilities based on visual input is a remarkable trait of the human species, and much of how exactly we do what we do—and seem to do it so well—remains to be discovered. Practical Image and Video Processing Using MATLAB®. By Oge Marques. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc. 3 4 INTRODUCTION AND OVERVIEW The need to extract information from images and interpret their contents has been one of the driving factors in the development of image processing1 and computer vision during the past decades. Image processing applications cover a wide range of human activities, such as the following: Medical Applications: Diagnostic imaging modalities such as digital radiogra- phy, PET (positron emission tomography), CAT (computerized axial tomogra- phy), MRI (magnetic resonance imaging), and fMRI (functional magnetic res- onance imaging), among others, have been adopted by the medical community on a large scale. Industrial Applications: Image processing systems have been successfully used in manufacturing systems for many tasks, such as safety systems, quality control, and control of automated guided vehicles (AGVs). Military Applications: Some of the most challenging and performance-critical scenarios for image processing solutions have been developed for military needs, ranging from detection of soldiers or vehicles to missile guidance and object recognition and reconnaissance tasks using unmanned aerial vehicles (UAVs). In addition, military applications often require the use of different imaging sen- sors, such as range cameras and thermographic forward-looking infrared (FLIR) cameras. Law Enforcement and Security: Surveillance applications have become one of the most intensely researched areas within the video processing community. Biometric techniques (e.g., fingerprint, face, iris, and hand recognition), which have been the subject of image processing research for more than a decade, have recently become commercially available. Consumer Electronics: Digital cameras and camcorders, with sophisticated built-in processing capabilities, have rendered film and analog tape technolo- gies obsolete. Software packages to enhance, edit, organize, and publish images and videos have grown in sophistication while keeping a user-friendly inter- face. High-definition TVs, monitors, DVD players, and personal video recorders (PVRs) are becoming increasingly popular and affordable. Image and video have also successfully made the leap to other devices, such as personal digital assis- tants (PDAs), cell phones, and portable music (MP3) players. The Internet, Particularly the World Wide Web: There is a huge amount of visual information available on the Web. Collaborative image and video uploading, sharing, and annotation (tagging) have become increasingly popular. Finding and retrieving images and videos on the Web based on their contents remains an open research challenge. 1 From this point on, the use of the phrase image processing should be interpreted as digital image process- ing. We shall only use the digital qualifier when it becomes relevant (e.g., after an analog image has been converted to a digital representation). BASIC CONCEPTS AND TERMINOLOGY 5 1.2 BASIC CONCEPTS AND TERMINOLOGY In this section, we define the most frequently used terms in Part I of this book. Although there is no universal agreement on the terminology used in this field, the definitions presented here are consistently used throughout the book. This section is structured in a question-and-answer format. What Is an Image? An image is a visual representation of an object, a person, or a scene produced by an optical device such as a mirror, a lens, or a camera. This representation is two dimensional (2D), although it corresponds to one of the infinitely many projections of a real-world, three-dimensional (3D) object or scene. What Is a Digital Image? A digital image is a representation of a two-dimensional image using a finite number of points, usually referred to as picture elements, pels, or pixels. Each pixel is represented by one or more numerical values: for monochrome (grayscale) images, a single value representing the intensity of the pixel (usually in a [0, 255] range) is enough; for color images, three values (e.g., representing the amount of red (R), green (G), and blue (B)) are usually required. Alternative ways of representing color images, such as the indexed color image representation, are described in Chapter 2. What Is Digital Image Processing? Digital image processing can be defined as the science of modifying digital images by means of a digital computer. Since both the images and the computers that process them are digital in nature, we will focus exclusively on digital image processing in this book. The changes that take place in the images are usually performed automatically and rely on carefully designed algorithms. This is in clear contrast with another scenario, such as touching up a photo using an airbrush tool in a photo editing software, in which images are processed manually and the success of the task depends on human ability and dexterity. We refer to the latter as image manipulation to make this distinction more explicit. What Is the Scope of Image Processing? In this book, we adopt the terminology used in [GW08] (among others) and employ the term image processing to refer to all the techniques and applications described in Part I of this book, whether the output is a modified (i.e., processed) version of the input image, an encoded version of its main attributes, or a nonpictorial description of its contents. 6 INTRODUCTION AND OVERVIEW Moreover, we distinguish among three levels of image processing opera- tions [GW08]: Low Level: Primitive operations (e.g., noise reduction, contrast enhancement, etc.) where both the input and the output are images. Mid Level: Extraction of attributes (e.g., edges, contours, regions, etc.) from images. High Level: Analysis and interpretation of the contents of a scene. This book does not cover the area of computer graphics or image synthesis, the process by which a 2D or 3D image is rendered from numerical data. In fact, we are often interested in the opposite process (sometimes referred to as image analysis), by which textual and numerical data can be extracted from an array of pixels. Image processing is a multidisciplinary field, with contributions from different branches of science (particularly mathematics, physics, and computer science) and computer, optical, and electrical engineering. Moreover, it overlaps other areas such as pattern recognition, machine learning, artificial intelligence, and human vision research. This combination of cross-disciplinary research and intersecting fields can be seen in the list of magazines and journals presented in Section 1.6. 1.3 EXAMPLES OF TYPICAL IMAGE PROCESSING OPERATIONS Image processing covers a wide and diverse array of techniques and algorithms, which will be described in detail in the remainder of Part I of this book. In this section, we provide a preview of the most representative image processing operations that you will learn about in forthcoming chapters. 1. Sharpening (Figure 1.1): A technique by which the edges and fine details of an image are enhanced for human viewing. Chapters 8–10 will discuss how this is done in the spatial domain, whereas Chapter 11 will extend the discussion to frequency-domain techniques. 2. Noise Removal (Figure 1.2): Image processing filters can be used to reduce the amount of noise in an image before processing it any further. Depending on the type of noise, different noise removal techniques are used, as we will learn in Chapter 12. 3. Deblurring (Figure 1.3): An image may appear blurred for many reasons, rang- ing from improper focusing of the lens to an insufficient shutter speed for a fast-moving object. In Chapter 12, we will look at image deblurring algorithms. 4. Edge Extraction (Figure 1.4): Extracting edges from an image is a fundamental preprocessing step used to separate objects from one another before identify- ing their contents. Edge detection algorithms and techniques are discussed in Chapter 14. EXAMPLES OF TYPICAL IMAGE PROCESSING OPERATIONS 7 FIGURE 1.1 Image sharpening: (a) original image; (b) after sharpening. 5. Binarization (Figure 1.5): In many image analysis applications, it is often nec- essary to reduce the number of gray levels in a monochrome image to simplify and speed up its interpretation. Reducing a grayscale image to only two levels of gray (black and white) is usually referred to as binarization, a process that will be discussed in more detail in Chapter 15. FIGURE 1.2 Noise removal: (a) original (noisy) image; (b) after removing noise. 8 INTRODUCTION AND OVERVIEW FIGURE 1.3 Deblurring: (a) original (blurry) image; (b) after removing the (motion) blur. Original image: courtesy of MathWorks. 6. Blurring (Figure 1.6): It is sometimes necessary to blur an image in order to minimize the importance of texture and fine detail in a scene, for instance, in cases where objects can be better recognized by their shape. Blurring tech- niques in spatial and frequency domain will be discussed in Chapters 10 and 11, respectively. 7. Contrast Enhancement (Figure 1.7): In order to improve an image for human viewing as well as make other image processing tasks (e.g., edge extraction) FIGURE 1.4 Edge extraction: (a) original image; (b) after extracting its most relevant edges. Original image: courtesy of MathWorks. EXAMPLES OF TYPICAL IMAGE PROCESSING OPERATIONS 9 FIGURE 1.5 Binarization: (a) original grayscale image; (b) after conversion to a black-and- white version. Original image: courtesy of MathWorks. FIGURE 1.6 Blurring: (a) original image; (b) after blurring to remove unnecessary details. Original image: courtesy of MathWorks. FIGURE 1.7 Contrast enhancement: (a) original image; (b) after histogram equalization to improve contrast. 10 INTRODUCTION AND OVERVIEW FIGURE 1.8 Object segmentation and labeling: (a) original image; (b) after segmenting and labeling individual objects. Original image: courtesy of MathWorks. easier, it is often necessary to enhance the contrast of an image. Contrast en- hancement techniques using transformation functions and histogram processing will be discussed in Chapters 8 and 9, respectively. 8. Object Segmentation and Labeling (Figure 1.8): The task of segmenting and labeling objects within a scene is a prerequisite for most object recognition and classification systems. Once the relevant objects have been segmented and labeled, their relevant features can be extracted and used to classify, compare, cluster, or recognize the objects in question. Segmentation and labeling of con- nected components from an image will be discussed in Chapters 13 and 15. Feature extraction and representation, and pattern recognition will be covered in Chapters 18 and 19, respectively. 1.4 COMPONENTS OF A DIGITAL IMAGE PROCESSING SYSTEM In this section, we present a generic digital image processing system and discuss its main building blocks (Figure 1.9). The system is built around a computer in which most image processing tasks are carried out, but also includes hardware and software for image acquisition, storage, and display. The actual hardware associated with each block in Figure 1.9 changes as technology evolves. In fact, even contemporary digital still cameras can be modeled according to that diagram: the CCD sensor corresponds to the Acquisition block, flash memory is used for storage, a small LCD monitor for display, and the digital signal processor (DSP) chip becomes the ‘Computer’, where certain image processing operations (e.g., conversion from RAW format to JPEG2 ) take place.