Automatic Skin Lesion Segmentation using SegNet, a Deep Learning architecture with certain extra requirements. Keeping the pre- and post-processing of the photos to a minimum is the secondary goal. The dermoscopic pictures in the PH2 dataset, which are part of the limited amount of images used to train the proposed model, were manually segmented.