Apurba Das

Image Enhancement in Spatial Domain

Image enhancement is a process, rather a preprocessing step, through which an original image is made suitable for a specific application. The application may vary from thermal image to X-ray image and accordingly the process of image enhancement would differ. Generally, the effect of image enhancement can be perceived visually. Even to address/handle the regular artifacts due to geometric transformations of images, image enhancement is done in form of image interpolation. The spatial domain refers to the 2D image plane represented in terms of pixel intensities. When the image is enhanced by modifying the pixel intensities directly (not as an effect of some other parameter tuning in a different domain), the method is considered as spatial domain image enhancement methodology. Otherwise, the image can be transformed to some other domain—like one 2D image can be transformed to a 2D frequency domain by discrete Fourier transform (DFT). To achieve an enhanced image, the Fourier coefficients are modified. That family of image enhancement methodologies is considered as frequency domain image enhancement which is discussed in the subsequent chapters. Whatever be the domain of image enhancement (either spatial or frequency domain), by the term image enhancement we mean improvement of the appearance of an image (in all sense including human perception and machine perception) by increasing the dominance of some features, or by decreasing the ambiguity between different regions of the image. In most cases, the enhancement of certain features is achieved at the cost of suppressing few other features.

Hybridization of DCTune and psycho-visual saliency model to improve security and capacity in DCT based image watermarking

Watermarking in images is necessary for authentication of digital library based applications, mostly. For efficient utilization of storage space and bandwidth of communication channel, compressed images ensuring fair amount of quality is employed. In our method we have addressed compressed images and found relatively redundant DCT coefficients for data embedding with maximized capacity. Here the concept of psycho-visual saliency and DCTune is combined to judge the capacity of each region of an image and the interpretation of saliency in frequency domain is exploited to ensure the watermarking technique adaptive.

PAA: Perception based Anti-Aliasing for High Quality Transformed Image Reconstruction from Single Image

Multiple occurrences of single object in same page or different pages of same document can generally be handled by a repository of reusable document component (RDCR). The same objects of different size and orientation could also be tagged in the same RDCR when orientation and scale maps are used as pointers. Hence after object collection and flattening in document image processing system, the compression ratio would be improved by employing OS-RDC (Orientation and Scale Mapped Reusable Document Component) which in turn would improve the performance of the system (as printing system). But especially for the scaled down and non-orthogonally rotated objects, jagging effect creates significant imaging artifacts in terms of image quality degradation. The current paper addresses to improve the perceptual image quality through OS-RDC and perception based anti-aliasing (PAA) filter. The combination of OS-RDC (used to achieve better speed) and PAA (used to improve the quality of the geometric transformed images) ensures both compression ratio and IQ (Image Quality) which are generally antagonistic in nature.

Patterns in Images and Their Applications

“Recognition” is a very important task of any intelligent system. When we are particularly interested in enhancing the ability of a computer by incorporating a vision system in it, in other words the capability of perceiving an image and processing the same, the recognition becomes really important in order to declare a system to be intelligent.

Color Science and Color Technology

We cannot see energy, we just can feel the effect of it. Light is also an energy and we can only see the objects after light falls on it and the reflected light reaches our eyes and psycho-visual system from the target object. The perception of color hence is dependent on light, reflectance spectra of object, and observer. The object property plays a very important role in color perception. Different devices responsible for color realization work in two major ways and the designation of primary colors also varies according to the system property (e.g., additive color and subtractive color model dependent on devices). In the current chapter, the visible electromagnetic spectra have been studied and primary colors are identified for additive color space. Next, depending on the color geometry, the primaries are derived for subtractive color space, too. The current chapter then shows how from reflectance spectra colors can be defined and derived. The psycho-visual model of color perception and its influence in derivation of device independent color space is discussed. Minute color measurement in order to ensure highest Color Image Quality (CIQ) is discussed in detail. In the color technology module, the concept of halftoning through error diffusion is presented along with gamut mapping with respect to different suitable rending intents.

Appendix A: Digital Differentiation and Edge Detection

This book has the prime objective to interpret an image in the form of both as a two-dimensional (2D) signal and as a pattern. In our introductory chapter, we have discussed the formation of image in nature and its digital coding for the ease of processing, communication, and storage. In this chapter, we physically interpret the method of convolution in the image-processing framework by the method of digital differentiation. Here, we take an example of edge detection from the digital image-processing domain to understand the convolution in light of digital differentiation.

Introduction to Digital Image

The word “signal” carries a broad meaning in all the domains of knowledge gathering, ranging from electronics and computer engineering to deaf-and-dumb communication. The word is uttered by every professional ranging from journalists to linguists. In the present chapter, we have introduced signal and tried to make the definition generic. From the basic definition of image, we will try to fit image into the generic definition of signal. As the title of the book signifies, here we will address the subject digital image processing through the two sided merely correlated guides to signal processing and pattern recognition.

Morphology-Based Image Processing

The word morphology refers to a discipline in biology where the shapes and structures of animals and trees are discussed and analyzed. In this chapter, we concentrate on the mathematics of morphology and apply them to image processing, especially binary images. To be precise, morphology-based image processing can be considered as the bridge between the signal and pattern property of image. This can be considered as both preprocessing algorithm and feature-extraction algorithm. Morphology-based image processing is specially used for shape-based feature extraction like boundary, convex, skeleton, and regions from an image.

Appendix C: Frequently Used MATLAB Functions

The frequently used Matlab functions, throughout this book, are defined in this appendix with examples.

Psycho-visual pattern recognition: Computer Vision

Object recognition is one of the most crucial and yet least understood aspects of visual perception. A simple answer to what we mean by visually recognizing an object may be, naming the object in sight. Humphreys et al. have identified several stages in visual processing that results in naming the object through recognition. However, while discussing this issue in detail, Ullman has shown that a natural association between naming and recognizing may not be all that unambiguous. This is because, an object may simultaneously belong to a number of classes like, for example, a book, my book, a comics book, Tintin in Tibet, and so on. From this example, it is clear that naming the object in sight depends upon the subjective recognition of the appropriate class as well, which in turn depends on the purpose of recognition under the given circumstances. Furthermore, even animals that cannot express themselves through language can still visually recognize objects. Humphreys et al. in their work have also acknowledged this issue of associated subjectivity in object recognition by making distinction among semantic representation, name representation, and semantic classification in their computational model that starts from a structural representation of the object. Significantly though, they have also demonstrated that a top-down intervention from semantic units to structural description units plays an important role in object naming in terms of top-down influence from higher to lower level in recognition of vision. However, in this chapter, our focus is on understanding the complex process of object recognition at the mid-level vision in terms of the several interacting components that are involved in it, especially the factors like intensity, orientation, and relative size of the region of attention.