Detection of Brain Tumor Using Zernike Moments on Magnetic Resonance Images

摘要

In this study, a new method has been developed for the detection of brain tumor in magnetic resonance images. Magnetic resonance images are analyzed using the Zernike moments and different orders of Zernike moments are calculated. The image is divided into two parts from the center of the image. The average value of the pixels located at the central line is calculated. The new vectors of the pixel are formed based on the calculated average value. The value obtained from the low and high order of Zernike moments are used to calculate the proper threshold value which can extract the tumor efficiently. The proposed method was tested with the different magnetic resonance images containing tumor and the algorithm was successful to analyze the tumor part from the brain image.