Improved fuzzy logic based edge detection method on clinical images

Abstract

Signal processing is the main field combining electrical engineering and mathematics, used to analyze digital and analog signals. Signal processing deals with the storage, compression, filtering and other processing of signals. These signals can be sound signals, image signals, and other signals. Nowadays images are essential thing for many area. Images can be used in space researches, military applications, marine workings, automotive industry, environment, agriculture and medical science. The area where the signal type is processed is that the input is an image, and the output is also an image, which is called image processing. Image Processing is one of the main research area in the disciplines of computer science and engineering. Image processing is a methods which performs operations on an image, on account of get an information from image. The progress of image processing are improved by the help of: the development of technology, the development of discrete theory, the demand for a pretty wide range of applications. It can be divided into digital and analog image processing. Image processing for analog images is used for hard copies of photos. Digital image processing uses computers to process digital images. Image processing has various kind of application such as sharpening, blurring, contrast adjustment, and edge detection etc. Edge detection is helpful for applications in the fields such as fingerprint matching, medical diagnosis, license plate detection, biomedical imaging, pattern recognition and machine vision. Edge detection technique makes the high intensity valued pixels visible. Edge detection is a compelling assignment. When edge detection must be applied to noisy images, it becomes more difficult. The idea of fuzzy logic helps to get rid of this problem with expert knowledge. The concept of fuzzy logic was first proposed in the 1960s by Professor Lütfi Aliasker Zade in Berkeley. Lütfi Aliasker Zade is committed to translating natural language into computer language, but it is not easy to translate into computer language terms 0 and 1. Zade proposed a shape of polyvalent logic within which the truth valuation of variables is also any real number between 0 and 1 whereas classical logic theory is utilizing with values false or true. Fuzzy logic can be summarized as predicated on the observation that individuals make decisions supported vague and non-numerical information. Fuzzy models are numerical implies of speaking to dubiousness and uncertain data. These models have the inclination of deciphering and controlling information and information that are non-certain. Additionally, it's conceivable to characterize linguistic variables like brief, exceptionally brief, long, or exceptionally long with fuzzy logic. Lütfi Zade's proposed theory fuzzy logic has been applied to various fields such as robotics, artificial intelligence, modeling and controlling system which is nonlinear or digital image processing. These fields used type-1 fuzzy logic until Prof. Lütfi A. Zade presented type-2 fuzzy logic in 1975. Fuzzy logic's type-2 theory was improved for uncertainties and non-linearity due on type-1 fuzzy rules, it shows fuzzy logic frameworks on type-2 are more fruitful than fuzzy logic frameworks on type-1 to unravel vulnerabilities. Be that as it may, working with fuzzy logic frameworks on type-2 are distant more advanced than working with fuzzy logic frameworks on type-1. In this thesis we will talk about a type-1 edge detection with fuzzy logic implementation for medical brain images, with the assistance of digital image, and digital image processing. This thesis gives you the performance comparison of widely used edge detection methods and improved edge detection with fuzzy logic method with interpreting digital images with the help of image enhancement and restoration and performing operations on images such as blurring, contrast adjustment. Different sources of digital images will be tested and results for each source will be provided.