Image Segmentation Algorithm for Semantic Segmentation with Sharp Boundaries using Image Processing and Deep Neural Network

摘要

Deep neural network (DNN)-based image segmentation has the endemic problems caused by iterative sampling, i.e., inaccurate and uncertain boundaries. On the other hand, conventional image segmentation based on image processing tends to extract well-aligned object boundaries, but has been difficult in semantic segmentation. In this paper, we propose an image segmentation method that exploits both image processing techniques and a DNN to extract semantic objects with well-aligned boundaries. In the experiments, it is confirmed that the proposed algorithm improves DNN-based segmentation effectively.