Skin lesion segmentation method for dermoscopic images with convolutional neural networks and semantic segmentation

Melanoma skin cancer is one of the most dangerous forms because it grows fast and causes deaths. Hence, early detection a very important task to treat melanoma. In this article, we propose lesion segmentation method for dermoscopic images based on U-Net architecture with VGG-16 encoder semantic segmentation. Base segmented lesion, diagnostic imaging systems can evaluate features classify them. The proposed requires fewer resources training, suitable computing without powerful GPUs, but training accuracy still high enough (above 95 %). experiments, train model ISIC dataset – common image dataset. To assess performance method, Sorensen-Dice Jaccard scores compare other deep learning-based methods. Experimental results showed that quality are better than ones compared