Preliminary detection of breast cancer (BC) is a major issue globally, and it needs to be addressed to help reduce fatalities of patients. Medical imaging in the form of mammogram scans is used for early breast cancer detection reducing costs incurred during treatment. This early detection of breast cancer lesions depends on various segmentation methods. Image segmentation is key for image analysis and includes detection, feature extraction, classification, and treatment. It assists physicians in measuring breast tissue to facilitate treatment. Segmentation of breast cancer is not easy due to the lesion’s geographical locations, complex tissue structures, and poor image quality caused by low contrast, noise, and inconsistent border edges. This paper proposes a BC histology image segmentation method using connected component analysis. Images in this method are pre-processed through data augmentation, gray scaling, and morphological operations such as dilation to remove noise and corrupted pixels. Consequently, BC image segmentation is done by connected components to identify and extract breast cancer lesions. This method evaluates an augmented dataset of 11,151 breast cancer histology images. It is trained on 8079 training set images and tested on 2694 testing set images. This proposed method yielded an outcome of \(98\%\) F1-accuracy score.