Abstract
Digital image processing is becoming another ever-growing yet popular field of demands for daily life, ranging from medicine, room evaluation, security, support, and security of the automotive community, among many others. The proposed framework focuses mostly on fuzzy logic structures somewhere in optical image processing. The main goal of most of this work is to demonstrate how fuzzy logic is implemented in image processing with little more than a quick introduction of fuzzy logic and optical image processing. Fuzzy logic, one of those artificial intelligence decision-making approaches, provides even greater room for use. When everything that would also have been allowed access to declarations at all since birth, particularly concerning in popularity in recent years, fuzzy logic as a whole has been proven to be true in virtually all systemic fields. Furthermore, the implications continue to suggest that the previously presented technique is worthy of attention in image processing software systems with the appropriate expansion.
TopIntroduction
In 1965 Lotfi A. Zadeh and Dieter Klaus developed Fuzzy sets independently as an enhancement of both the traditional instance of collection. Fuzzy relations are increasingly employed in a range of areas, including language studies, decision-making, and clustering. Fuzzy logic has been extended to a wide range of disciplines ranging from control theory to artificial intelligence. Fuzzy relations were designed to enable the machine to decide the distinctions that are neither right nor wrong between data.
The clustering algorithm separates a data collection into classes/clusters, with related data objects allocated to the same clusters. When the boundary between clusters is ill-defined, resulting in circumstances where the same data object belongs to more than one class, the concept of fuzzy clustering comes into play.
Clustering or cluster analysis have the steps like handing over of data points to clusters such that items in the same cluster are as analogous as possible, otherwise clusters are contradictory as possible.
Key Terms in this Chapter
Membership Function: Membership function shows the degrees of belongingness of elements of universe of discourse say X to set A where . Simply put, it is a function that maps some set of real numbers to the interval [0, 1]. i.e. ?? A : X®[0,1].
Unsupervised Learning: By applying machine learning algorithms to the analysis and classification of unlabeled datasets, unsupervised learning (also known as unsupervised machine learning) is able to shed light on previously opaque data. Without the help of a human, these algorithms can unearth previously unseen patterns or groups of related data.
Fuzzy Numbers: In contrast to a traditional real number, which can only ever refer to one specific value, a fuzzy number can refer to an infinitely large collection of related values, each of which is assigned a relative importance between zero and one. The concept of fuzzy numbers is a logical expansion of the classical number system.
Centroid: The location that represents the cluster's centre, whether it is real or imagined, is called a centroid. Each data point is then placed in a distinct cluster by reducing the in-cluster sum of squares.
Image Processing: Image processing is a technique for altering or extracting information from an image by the application of various mathematical, statistical, and computational methods. It's a form of signal processing where the input is an image and the output could be the same image or some of the image's features.
Fuzzy Sets: In 1965, Lotfi A. Zadeh introduced fuzzy sets. Fuzzy sets are those sets whose elements have a degree of membership i.e., every fuzzy set is characterized by a membership function.
Outliers: Since the majority of clustering methods require a minimum number of data points to construct a cluster, outliers are the ungrouped data points. Even if the outliers form a cluster, it is quite isolated from other clusters.
Noise: Data points that are anomalous or do not form a part of any significant cluster are referred to as noise.