BRAIN TUMOR IMAGE SEGMENTATION USING INTELLIGENT MEAN SHIFT CLUSTERING TECHNIQUE
Journal Title: Elysium Journal of Engineering Research and Management - Year 2015, Vol 2, Issue 3
Abstract
Brain tumor is a deadly disease which challenges on detecting tumor cells. The tumor detection becomes more complicated for diagnosis as it exhibits complex characteristics. To address this problem, we propose a brain tumor segmentation method for identification of tumor cells by iterative process. Mean shift clustering technique combines with local independent projection – based classification (LIPC) to detect tumor by means of iterative process. We introduce statistical regional merging (SRM) the change does not require prior knowledge of the means of the average number of clusters and the clusters are not controlling in the form, it is a nonparametric clustering technique. It is used for locating the maxima of a density function. Local independent project - based classification (LIPC), where all the values applied to the closest match. Once the tumor cells and normal cells Bhattacharyya coefficient as a result of the segmentation is to improve the classification performance. The test results for several categories of data are evaluated in the global brain images and a common platform. The additional advantage of the proposed method is to apply the universal brain images and a common platform.
Authors and Affiliations
Priyanka V, Kathirvel K. , Batri K.
Keywords
Related Articles
- EP ID EP392121
- DOI -
- Views 100
- Downloads 0
DETECTION AND PREVENTION OF MALICIOUS FEEDBACK RATING IN WEB SERVICE RECOMMENDATION SYSTEMS
To facilitate a recommendation to users for selecting an appropriate or reputed web service from mass of services by avoiding dishonest recommendation. To prevent, avoid and de...
AN EFFICIENT ALGORITHM FOR ESTIMATING THE HISTOGRAM USING SPATIO-TEMPORAL DATA
This paper spotlight on an influential query in methodical reproduction data analysis: the histogram estimation on spatial data. Histogram is nothing but an split up of datas into buckets for optimizing t...
PREDICTION IN DATA MINING COMPARATIVE STUDY OF HEART DISEASE
Data mining is the process of finding useful and relevant information from the databases. There are several types of data mining techniques are available. Association Rule, Classification, Neur...
REAL TIME SECTIONALIZATION OF ENHANCED SHARPNESS VIDEO USING FPGA
Background Identification is a general feature in many video privilege systems. Gaussian Mixture Models (GMM) is one of the popular fashionable and winning approaches to complete Background identification...
A SURVEY ON LUNG SEGMENTATION TECHNIQUES
The cancer in the lung is the common cancer and leads to dead often. The cells which are affected are difficult to analyze in the early stage of cancer because they are overlapped. Many techniques are u...