1
Introduction

A quantum computer, as the name suggests, fundamentally works on several quantum physical characteristics. It is also considered as the field of study, primarily focused on evolving computer technology using the features of quantum theory, which expounds the nature of energy and substance and its behavior on the quantum level, i.e., at the atomic and subatomic level. Developing a quantum computer would mark an advance in computing competency far superior to any current computers. Thus, the use of quantum computers could be an immense improvement on current computers because they have enormous processing capability, even exponentially, compared to classical computers. The supremacy of processing is gained through the capacity of handling multiple states at once, and performing tasks exploiting all the promising permutations in chorus. The term quantum computing is fundamentally a synergistic combination of thoughts from quantum physics, classical information theory, and computer science.

Soft computing (SC), introduced by Lotfi A. Zadeh [282], manages the soft meaning of thoughts. SC, comprising a variety of thoughts and practices, is fundamentally used to solve the difficulties stumbled upon in real-life problems. This can be used to exploit the uncertainty problem almost with zero difficulty. This can also handle real-world state of affairs and afford lower solution costs [29]. The advantageous features of SC can best be described as leniency of approximation, vagueness, robustness, and partial truth [103,215]. This is a comparatively novel computing paradigm which involves a synergistic amalgamation of essentially several additional computing paradigms, which may include fuzzy logic, evolutionary computation, neural networks, machine learning, support vector machines, and also probabilistic reasoning. SC can combine the aforementioned computing paradigms to offer a framework for designing many information processing applications that can function in the real world. This synergism was called computational intelligence by Bezdek [24]. These SC components are different from each other in more than one way. These can be used to operate either autonomously or conjointly, depending on the application domain.

Evolutionary computation (EC) is a search and optimization procedure which uses biological evolution inspired by Darwinian principles [14,83,136]. It is stochastic and delivers robust search and optimization methods. It starts with a pool of trial solutions in its search space, which is called the population. Numerous in-built operators are generally applied to each individual of the population, which may cause population diversity and also leads to better solutions. A metric, called the fitness function (objective function), is employed to determine the suitability of an individual in the population at any particular generation. As soon as the fitness of the existing individuals in the population is computed, the operators are successively applied to produce a new population for the successive generations. Distinct examples of EC may include the Differential Evolution [242], Genetic Algorithms [127,210], Particle Swarm Optimization [144], and Ant Colony Optimization [196], to name but a few. Simulated annealing [147] is another popular example of meta-heuristic and optimization techniques in this regard. This technique exploits the features of statistical mechanics concerning the behavior of atoms at very low temperature to find minimal cost solutions of any given optimization problem. EC techniques are also useful when dealing with several conflicting objectives, called the multi-objective evolutionary techniques. These search procedures provide a set of solutions, called optimal solutions. Some typical examples of these techniques may include the multi-objective differential evolutionary algorithm (MODE) [275], the multi-objective genetic algorithm (MOGA) [172,183], and multi-objective simulated annealing (MOSA) [237], to name but a few.

Fuzzy logic tenders more elegant alternatives to conventional (Boolean) logic. Fuzzy logic is able to handle the notion of partial truth competently [139,141,215,282,283]. A neural network is a computing framework comprising huge numbers of simple, exceedingly unified processing elements called artificial neurons, which add up to an elemental computing primitive [82,102,150]. Machine learning is a kind of intelligent program which works on example data. It learns from previous experiences and is used to enhance the performances by optimizing a given criterion [5,156,178]. Support vector machines (SVM) are known to be the collection of supervised learning techniques. SVMs are very useful in regression and classification analysis [38,50]. SVMs are fit to handle a number of real-life applications, including text and image classification, or biosequences analysis, to name but a few [38,50]. Nowadays SVMs are often used as the standard and effective tool for data mining and machine learning activities. Probabilistic reasoning can be defined as the computational method which uses certain logic and probability theory to handle uncertain circumstances [201,202].

Many researchers utilize the basic features of quantum computing in various evolutionary algorithmic frameworks in the soft computing discipline. The underlying principles of quantum computing are injected into different meta-heuristic structures to develop different quantum inspired techniques. In the context of image analysis, the features are extracted both from pictographic and non-numeric data and are used in these algorithms in different ways [27]. This chapter provides an insight into the various facets of the quantum computing, image segmentation, image thresholding, and optimization. This chapter is arranged into a number of relevant sections. Section 1.1 presents an overview of the underlying concepts of image analysis. A brief overview of image segmentation and image thresholding is discussed in this section. Section 1.2 throws light on the basics of quantum computing in detail. Section 1.3 discusses the necessity of optimization in the real world. This section presents different types of optimization procedures with their application in the real world. Apart from the above issues, this chapter also presents a short description of the literature survey on related topics. Different types of approaches in this regard are in detail presented in Section 1.4. The organization of the book is presented in Section 1.5. The chapter concludes in Section 1.6. It also shows the direction of research that can be used as future reference. A brief summary of the chapter is given in Section 1.7. In Section 1.8, a set of questions related to the theme of the chapter is presented.

1.1 Image Analysis

Image analysis has a vital role in extracting relevant and meaningful information from images. There are few automatic or semi-automatic techniques, called computer/machine vision, pattern recognition, image description, image understanding to name but a few, used for this purpose. Image segmentation can be thought of as the most fundamental and significant step in several image analysis techniques. A good example of image analysis may involve the organized activities of the human eye with the brain. Computer-based image analysis can be thought of as the best alternative which may reduce human effort in order to make this process faster, more efficient, and automatic. Image analysis has numerous applications in a variety of fields such as medicine, biology, robotics, remote sensing, and manufacturing. It also makes a significant contribution in different industrial activities such as process control, quality control, etc. For example, in the food industry, image analysis plays a significant role to ensure the uniform shape, size and texture of the final food products.

In medical image analysis, clinical images of different views are captured to diagnose and detect diseases in relation to body organs, and study standard physiological procedures for future references. These investigations can be accomplished through images attained from various imaging technologies, such as magnetic resonance imaging (MRI), radiology, ultrasound, etc. For example, image analysis methodology is of the utmost importance in cancer detection and diagnosis [44], thus it helps the physician to ensure accurate treatment for their patient. In the context of cancer treatment, several features like shape, size, and homogeneity of a tumor are taken into consideration when classifying and diagnosing cancer images. Different image analysis algorithms can be introduced that can help radiologists to classify tumor images.

Figure 1.1 Steps in image analysis.

The steps involved in image analysis are presented in Figure 1.1 [112]. Each step is discussed in the following in brief.

1. Image acquisition: This is the first step of every vision system. Image acquisition means acquiring a digital image. After obtaining the image successfully, several processing approaches can be used on the image in order to fulfill the different vision tasks required nowadays. However, if the image cannot be acquired competently, the anticipated tasks may perhaps not be completed successfully by any...