Brain tumor segmentation and classification using machine learning techniques

The goal of image segmentation is to partition an image to disjointed segments. This process is one of the most difficult tasks in image processing with a large practical impact. It is widely used for example in medical image processing for detecting veins, tissues, bones and other parts of the human body and their visualization or in computer vision for object tracking and many others. Moreover, many higher level algorithms rely on good results from segmentation algorithms. Although there are many approaches producing reasonable results, there is still a room for improvements, i.e. to speed the approaches up and to make them able to deal with worst and more complicated images, to make them be able to solve more complex and time critical issues like real-time object tracking or segmentation of damaged images. In this thesis we proposed an image segemention for brain tumor using classification methods with MRI images. Our method can be summerzied in four steps , first the feature extraction , the the feature selection , finaly the image segemention and image calssfication.