Detection and Classification of breast cancer in mammogram using textual and statistical features of image
Project involves extracting textual features of mammogram image using Grey-Level Cocurrence Matrix and classification of mammograms into Abnormal and Normal class using Random Forest classifier. Gaussian Filtering is incorporated for image enhancement and smoothing which reduces noise from image. Best GLCM features were selected by analyzing features scores obtained from AdaBoost classifier. Aim is to improve upon existing research work by trying different algorithms and extracting more powerful features.Overall accuracy achieved is 93.90%, which is comparable with most of the past research. Dataset used is MIAS.