摘要
近年来,深度卷积神经网络应用于图像语义分割领域并取得了巨大成功。提出了一个基于RGB-D(彩色-深度)图像的场景语义分割网络;该网络通过融合多级RGB网络特征图和深度图网络特征图,有效提高了卷积神经网络语义分割的准确率。同时,利用带孔的卷积核设计了具有捷径恒等连接的空间金字塔结构来提取高层次特征的多尺度信息。在SUN RGB-D数据集上的测试结果显示,与其他state-of-the-art的语义分割网络结构相比,所提出的场景语义分割网络性能突出。
In recent years,deep convolutional neural networks have been applied to the field of image semantic segmentation and achieved great success.A scene semantic segmentation network based on(Rredgreenblue-depth RGB-D) images was presented.The network effectively improves the accuracy of semantic segmentation of convolutional neural networks by merging multi-level RGB network features and depth network features.At the same time,convolution kernels with holes designs a spatial pyramid structure with shortcut to extract high-level features of multi-scale information was used.The test results on the SUN RGB-D dataset show that,compared with other stateof-the-art semantic segmentation networks,the performance of the semantic segmentation network proposed is outstanding.
作者
代具亭
汤心溢
刘鹏
邵保泰
DAI Ju-ting;TANG Xin-yi;LIU Peng;SHAO Bao-tai(Shanghai Institute of Technical Physics of the Chinese Academy of Sciences,Shanghai 200083,China;University of Chinese Academy of Sciences,Beijing 100084,China;Key Laboratory of Infrared System Detection and Imaging Technology,GAS,Shanghai 200083,China)
出处
《科学技术与工程》
北大核心
2018年第20期286-291,共6页
Science Technology and Engineering
基金
国家十三五国防预研项目(Jzx2016-0404/Y72-2)
中国科学院青年创新促进会(2014216)
上海市现场物证重点实验室基金(2017xcwzk08)资助
关键词
RGB-D
卷积神经网络
语义分割
特征融合
空间金字塔
RGB-D
convolutional neural networks
semantic segmentation
feature fusion
spatial pyramid