Brain Tumor Detection Using Automatic Region Growing Segmentation

Arti Bahuguna, Dr. H. S. Bhadauria

Abstract: Brain tumor detection and segmentation is very difficult task in MRI images. The MRI image generates a better contrast that indicates regular and irregular tissues. Its better for overlapping in margin of each limb. We use several automatic segmentation techniques. That may used to compensate with the problem if no growth of tumour and any small white part done. In this case when the tumor edge is not pointed then segmentation results are not accurate i. e. segmentation may be under or over. That procedure is done due to initial stage of the tumors. Segmentation of the Image is one of the ground laying techniques of digital image system.

Keywords: Segmentation of image, Automatic Region Growing, Unseeded point selection