Publication

Abstract : The most common cancer in women is breast cancer, which can be fatal if not detected in a timely manner. One of the newest fields in today's diagnostic system is digital pathology. The first crucial step before cancer classification is image segmentation. The fundamental problem occurs when there are nuclei for separation that are touching or overlapping. In this paper, an image segmentation approach is presented using the thresholding methodology. The morphological processes and binarization that follow OTSU thresholding are used in the proposed technique. The undesirable areas of the photograph are deleted during post-processing. The experimental results on the used datasets support the suggested strategy's effectiveness. The dataset will be examined as part of our second goal, which is to evaluate how well machine learning techniques may then be used to predict breast cancer. Here, Support Vector Machine is utilised for categorization malignant and benign cancer.