###
计算机系统应用:2020,29(9):156-163
本文二维码信息
码上扫一扫!
基于改进Mask-RCNN的遥感影像建筑物提取
何代毅1,2,3, 施文灶1,2,3, 林志斌1,2,3, 乔星星1,2,3, 刘芫汐1,2,3, 林耀辉1,2,3
(1.福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117;2.福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007;3.福建师范大学 福建省光子技术重点实验室, 福州 350007)
Building Extraction from Remote Sensing Image Based on Improved Mask-RCNN
(1.Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China;2.Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China;3.Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China)
摘要
图/表
参考文献
相似文献
本文已被:浏览 109次   下载 90
投稿时间:2019-12-04    修订日期:2019-12-27
中文摘要: 由于遥感影像中建筑物种类繁多且与周围环境信息混淆,传统方法难以实现建筑物的准确高效提取.本文提出了一种基于改进Mask-RCNN的建筑物自动提取方法,利用PyTorch深度学习框架搭建改进Mask-RCNN网络模型架构,在网络的设计中添加了路径聚合网络和特征增强功能,通过监督和迁移学习的方式对Inria航空影像标签数据集进行多线程迭代训练与模型优化学习,实现了建筑物的自动精确分割和提取.基于不同开源数据集,分别与SVM、FCN、U-net和Mask-RCNN等建筑物提取算法进行对比,实验表明,本文方法可以高效准确、高效地提取建筑物,对于同一个数据集,提取结果的mAPmRecallmPrecisionF1分数这4个评价指标均优于对比算法.
中文关键词: 建筑物提取  Mask-RCNN  PyTorch  实例分割
Abstract:In view of the variety of buildings, and the confusion with the surrounding environment in the remote sensing image, the traditional methods are difficult to extract the buildings efficiently and accurately. This study proposes a method of building automatic extraction method based on the improved Mask-RCNN. In the proposed method, an improved Mask-RCNN network model framework is constructed by using PyTorch deep learning framework, path aggregation network and feature enhancement function are added in the network design, and multi-threaded iterative training and model optimization learning are conducted for the Inria aerial image tag data set by means of supervision and migration learning, thus automatic and accurate segmentation and extraction of buildings are achieved. The proposed method is compared with SVM, FCN, U-net, Mask-RCNN, and other building extraction algorithms with different open-source datasets. The experimental results show that the proposed method has competitive performance, which can extract buildings more efficiently, accurately, and quickly in different open-source datasets, and the four evaluation indexes of mAP, mRecall, mPrecision, and F1 scores extracted in the same dataset are better than the compared algorithms.
文章编号:7484     中图分类号:    文献标志码:
基金项目:国家自然科学基金青年基金(41701491);福建省自然科学基金面上项目(2017J01464,2018J01619)
作者单位E-mail
何代毅 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
 
施文灶 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
swz@fjnu.edu.cn 
林志斌 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
 
乔星星 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
 
刘芫汐 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
 
林耀辉 福建师范大学 福建省光电传感应用工程技术研究中心, 福州 350117
福建师范大学 医学光电科学与技术教育部重点实验室, 福州 350007
福建师范大学 福建省光子技术重点实验室, 福州 350007 
 
Author NameAffiliationE-mail
HE Dai-Yi Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
 
SHI Wen-Zao Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
swz@fjnu.edu.cn 
LIN Zhi-Bin Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
 
QIAO Xing-Xing Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
 
LIU Yuan-Xi Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
 
LIN Yao-Hui Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application, Fujian Normal University, Fuzhou 350117, China
Key Laboratory of Optoelectronic Science and Technology for Medicine (Ministry of Education), Fujian Normal University, Fuzhou 350007, China
Fujian Provincial Key Laboratory of Photonics Technology, Fuzhou 350007, China 
 
引用文本:
何代毅,施文灶,林志斌,乔星星,刘芫汐,林耀辉.基于改进Mask-RCNN的遥感影像建筑物提取.计算机系统应用,2020,29(9):156-163
HE Dai-Yi,SHI Wen-Zao,LIN Zhi-Bin,QIAO Xing-Xing,LIU Yuan-Xi,LIN Yao-Hui.Building Extraction from Remote Sensing Image Based on Improved Mask-RCNN.COMPUTER SYSTEMS APPLICATIONS,2020,29(9):156-163

用微信扫一扫

用微信扫一扫