Finding the trouble spots on a breast mammogram is one of the hardest and most demanding things to do these days because of the changeable mass structure. Also, finding breast cancer early is important for getting the right care while the cancer is still in its earliest phases. A computer-aided diagnosis (CAD) model is usually designed for this purpose, but it only solves a few of the main issues, such as decreasing accuracy, taking a long time to train, making algorithms more complicated, and not being efficient when working with very large datasets. So, the point of this study is to come up with advanced segmentation and classification methods that will make it possible to accurately find breast cancer in mammogram images. The adaptive histogram equalisation enhancement (AHEE) model is first used to clean up the input picture by getting rid of noisy pixels and making the image better overall. Then, the colour dense model-based segmentation method is used to separate most of the areas with pixels that are very bright. Next, the set of characteristics from the picture that was successfully segmented is taken towards to make the classification more accurate. These features include deep, morphological, texture, and density. Finally, a classification method based on a convolutional neural network (CNN) is designed to discover if the received feature values from the image show a normal or abnormal state. A number of evaluation factors are also used to compare and confirm how well the suggested segmentation and classification process works. The colour dense model with CNN-based classification algorithm works better than the other methods by finding the problematic areas more accurately.