| Popis: |
In medical image processing, Brain Tumor segmentation plays a vital role to segment an image more accurately and precisely. Segmentation is the process used to accomplish the tasks by dividing an image into meaningful parts which share similar properties. Magnetic Resonance Imaging (MRI) is a primary diagnostic technique for image segmentation. It is challenging task due to poor contrast and artifact which results in missing or diffuse organ/tissue boundaries. This paper describes the curvature scale space algorithm for segmentation. It involves Pre- processing, Segmentation using Curvature Scale Space (CSS), Feature Extraction and Classification using ANN. The artificial neural network is used to train and classify the stages of Brain tumor as benign, malignant or normal. |
