This paper introduces a deep learning based image segmentation method. The method first uses a deep convolutional neural network to extract features, and then uses a segmentation network to divide the feature map into target and background. Experimental results show that the method achieves excellent performance on multiple image segmentation datasets.
