MUTLILEYER PERSEPTRON ANN BASED WATERSHED SEGMENTATION TO DETECT MAMMOGRAM TUMOR

ABSTRACT Image segmentation plays an essential role in many medical image processing and its helps to the physicians without experience in making right analyses. The watershed segmentation with Multi-layer perceptron-Artificial neural networks (MLP-ANN) are used in this paper for segmentation. The Morphological operations are applied, in order to remove noise and regional maxima that correspond to calcifications. The segmented objects are classified using neural networks. This work incorporates the benefits of neural networks characterization and it helps to detect and extract the tumor from the image. MLP is widely used for solving the problems and at the same time feature can extracted easily. The result can be evaluated with Activation function and quality measures. Keywords: Image segmentation, watershed algorithm, Multi-layer perceptron-Artificial neural networks.