Nayak et al., International Journal on Emerging Technologies 11(2): 840-844(2020) 840 International Journal on Emerging Technologies 11(2): 840-844(2020) ISSN No. (Print): 0975-8364 ISSN No. (Online): 2249-3255 Brain Tumor Detection and Extraction using Type-2 Fuzzy with Morphology Dillip Ranjan Nayak 1 , Neelamadhab Padhy 2 and Basanta Kumar Swain 3 1 Research scholar School of Computer Science & Engineering Gandhi Institute of Engineering & Technology, University Gunupur (Odisha), India. Assistant Professor, Department of Computer Science & Engineering, Govt. College of Engineering Kalahandi, Bhawanipatna (Odisha), India. 2 Associate Professor, School of Computer Science & Engineering, Gandhi Institute of Engineering &Technology, University Gunupur (Odisha), India. 3 Assistant Professor, Department of Computer Science & Engineering, Govt. College of Engineering Kalahandi, Bhawanipatna (Odisha), India. (Corresponding author: Dillip Ranjan Nayak) (Received 04 January 2020, Revised 03 March 2020, Accepted 05 March 2020) (Published by Research Trend, Website: www.researchtrend.net) ABSTRACT: Segmentation is the key process for the detection of the brain tumor. Different kinds of thresholding methods are used by researchers in segmentation process in order to detect the brain tumors. This research paper tackles detection of brain tumor by avoiding any threshold mechanism over the MRI scan images for detection and extraction of brain tumor in patients. The proposed method for tumor detection is based on Type-2 fuzzy and morphological operators which help to alleviate the herculean task of exact location identification of brain tumor. First, we have applied Hamacher T co-norm(S norm) with triangular function for initial enhancement and followed by INT function to enhance the tumor position for finding the exact location. Finally, Erode operation is applied to get the more accurate location of tumor position from medical point of view. The proposed method has produced the clear picture as well as identified the exact location of brain tumor despite of vagueness in the original images. Keywords: Brain Tumor, Clustering. Morphological Operators, MRI, Segmentation, Type-2 fuzzy, Thresholding. Abbreviations: MRI, magnetic resource imaging. I. INTRODUCTION A brain tumor is a kind of inner cell growth inside the brain. Brain tumors appear in different sizes and shapes. Hence, early detection of brain tumor and diagnosis is a difficult process. It is observed that Magnetic Resonance Imaging (MRI) with fuzzy is a significant technique for locating tumor position. Fuzzy Image segmentation [1] and watershed segmentation [2] are mainly used to locate the boundaries of brain tumor. An image can be partitioned into multiple regions by adopting the image segmentation process. Several researchers suggested various algorithms for image segmentation of the brain images [3-7]. It is hard to detect the exact location of tumor and volume of the tumor due to the presence of vagueness in brain tissue. Moreover, the brain tumor cells are high intensity in nature due to the presence of portentous fluid. Therefore, fuzzy function is the appropriate technique for classification of tumors. Now a day, there are many fuzzy computing methodologies which are used for classifying MR images [8-10]. Fuzzy clustering [11] and segmentation using Neural network [12] are also used to detect tumors from MR images. Intensity adjustment is also applied to detect and segment brain tumor [13]. Mainly, Type-1 fuzzy technique is used In fuzzy thresholding and fuzzy clustering scenario. But, finding exact thresholding point or clustering point is a difficult task in Type-1 fuzzy techniques. It is also found in the research that Type-1 fuzzy algorithms are demanding high computational time due to the uncertainty of information. However, our proposed Type-2 fuzzy sets with morphology helps to detect the exact brain tumor by combining the advantages of rough set and fuzzy set [14]. In this research article we have observed that, Type-2 fuzzy sets yielded improved performance as it can cope with high degree of uncertainties which is not possible in Type-1 fuzzy algorithm. It is also seen that Type-2 fuzzy set produces pretty good real-time response in brain tumor detection and extraction. Type-2 fuzzy set is an extension of two dimensional type-1 fuzzy set into the three dimensional. Type-2 fuzzy sets are highly used to get the true membership function for a fuzzy condition [15] in uncertainty scenario. The main motto of this research work is to get the information about the tumor using Hamacher T co- norm [16, 17] with fuzzy morphological operations. Authors have presented this research article in different sections as follows: Section II describes about material and methods adopted in the current work. Results are discussed in Section III. Section IV and Section V depict about conclusion and future work respectively. II. MATERIALS AND METHODS The uncertainties are mathematically modeled using Type-2 fuzzy sets. The basic theory of Type-2 fuzzy sets was introduced by Zadeh [18]. Image can be enhanced using Type-2 fuzzy set [19]. The truncated Type-2 fuzzy set and triangular Type-2 fuzzy logic are mostly used in getting brain tumor detection [20, 21]. Type-1 fuzzy set defined as A= {(x,   |x ϵ X} where    : X ∈ [0, 1] with the membership function on an element x varies as 0≤   ≤ 1. Chaira [16] defined e t