Recent progress in semantic image segmentation
Recent progress in semantic image segmentation
复制标题
语义图像分割的最新进展
DOI：
10.1007/s10462-018-9641-3
复制
发表时间：
2019-08-01
影响因子：
12
通讯作者：
Yang, Yuhan
中科院分区：
文献类型：
--
作者：
Liu, Xiaolong;Deng, Zhidong;Yang, Yuhan
语义图像分割是图像处理和计算机视觉领域的关键应用之一，已广泛应用于医疗、智能交通等多个领域。研究人员发布了大量的基准数据集来验证他们的算法。语义分割已经研究了很多年。自深度神经网络（Deep Neural Network， DNN）出现以来，分割技术取得了巨大的进步。本文将语义图像分割方法分为两类：传统的深度神经网络方法和最新的深度神经网络方法。本文首先对传统的分割方法和已发布的数据集进行了简要的总结，然后对近年来基于深度神经网络的分割方法进行了全面的研究，从八个方面进行了描述：全卷积网络、上样本方法、FCN与CRF联合方法、扩展卷积方法、骨干网络的进展、金字塔方法、多层次特征与多阶段方法、监督、弱监督和无监督方法。最后，对这一问题进行了总结。
Semantic image segmentation, which becomes one of the key applications in image processing and computer vision domain, has been used in multiple domains such as medical area and intelligent transportation. Lots of benchmark datasets are released for researchers to verify their algorithms. Semantic segmentation has been studied for many years. Since the emergence of Deep Neural Network (DNN), segmentation has made a tremendous progress. In this paper, we divide semantic image segmentation methods into two categories: traditional and recent DNN method. Firstly, we briefly summarize the traditional method as well as datasets released for segmentation, then we comprehensively investigate recent methods based on DNN which are described in the eight aspects: fully convolutional network, up-sample ways, FCN joint with CRF methods, dilated convolution approaches, progresses in backbone network, pyramid methods, Multi-level feature and multi-stage method, supervised, weakly-supervised and unsupervised methods. Finally, a conclusion in this area is drawn.