The proposed methodology utilizes the convolutional neural network (CNN) architecture U-Net for image segmentation and then applies the CNN architecture InceptionV3-Net for fault classification.