基于深层卷积残差网络的航拍图建筑物精确分割方法

doi:10.11896/jsjkx.200500096

计算机科学 ›› 2021, Vol. 48 ›› Issue (8): 169-174.doi: 10.11896/jsjkx.200500096

• 计算机图形学& 多媒体 • 上一篇下一篇

基于深层卷积残差网络的航拍图建筑物精确分割方法

许华杰^1,2, 张晨强¹, 苏国韶³

1 广西大学计算机与电子信息学院南宁530004
2 广西多媒体通信与网络技术重点实验室南宁530004
3 广西大学土木建筑工程学院南宁530004

收稿日期:2020-05-21 修回日期:2020-10-23 发布日期:2021-08-10
通讯作者: 许华杰(hjxu2009@163.com)
基金资助:
广西壮族自治区科技计划项目(2017AB15008);崇左市科技计划项目(FB2018001);广西高等学校高水平创新团队及卓越学者计划

Accurate Segmentation Method of Aerial Photography Buildings Based on Deep Convolutional Residual Network

XU Hua-jie^1,2, ZHANG Chen-qiang¹, SU Guo-shao³

1 College of Computer and Electronic Information,Guangxi University,Nanning 530004,China;
2 Guangxi Key Laboratory of Multimedia Communications and Network Technology,Nanning 530004,China;
3 College of Civil Engineering and Architecture,Guangxi University,Nanning 530004,China

Received:2020-05-21 Revised:2020-10-23 Published:2021-08-10
About author:XU Hua-jie,born in 1974,Ph.D,associa-te professor,is a senior member of China Computer Federation.His main research interests include artificial intelligence,acoustic signal recognition and computer vision.
Supported by:
Science and Technology Plan Project of Guangxi Zhuang Autonomous Region (2017AB15008),Science and Technology Plan Project of Chongzuo(FB2018001) and High Level Innovation Team and Outstanding Scholar Program of Universities in Guangxi Province.

摘要/Abstract

摘要： 针对建筑物3D建模场景下所需的建筑物主体轮廓俯视平面图获取成本较高、航拍图建筑物的分割精度低、建筑物屋顶存在干扰物影响分割等问题,文中提出了一种将5个点的位置表示为热图作为网络额外输入通道的基于深层残差网络的航拍图建筑物精确分割方法,该方法在航拍图建筑物的精确分割任务中取得了比较好的分割效果。实验结果表明,该方法具有比传统半自动分割方法Grabcut更高的分割精度和分割效率;具有比DEXTR方法更好的鲁棒性和抗干扰性。该方法可以为建筑物3D重建任务提供高精度的建筑物俯视轮廓图和建筑物顶部图片,还可以在航拍图建筑物数据集的制作过程中,作为一种准确和有效的掩码注释工具或半自动轮廓标注工具,以提高数据集的标注效率。

关键词: 3D建模, 航拍图, 卷积残差网络, 热图, 图像分割

Abstract: In order to solve the problems of high cost of obtaining the top plan view of the main outline of the building in the 3D modeling scenario,low segmentation accuracy of the aerial photography building,interference on the roof of the building,etc.,a method of accurately segmenting the aerial photography building based on deep residual network is proposed,in which the positions of five points are expressed as heat maps as additional input channels of the network,and good segmentation effect is achieved in the task of accurately segmenting the aerial photography building.Experimental results show that the proposedmethod has higher segmentation accuracy and segmentation efficiency than the traditional semi-automatic segmentation method Grabcut.It has better robustness and anti-interference than DEXTR method.This method can provide high-precision top-view contour map and top-view picture of buildings for 3D reconstruction of buildings,and can also be used in the production process of aerial photography building data sets as an accurate and effective mask annotation tool or semi-automatic contour annotation tool to improve the annotation efficiency of datasets.

Key words: 3D modeling, Aerial photography, Convolutional residual network, Heatmap, Image segmentation

中图分类号:

TP391

许华杰, 张晨强, 苏国韶. 基于深层卷积残差网络的航拍图建筑物精确分割方法[J]. 计算机科学, 2021, 48(8): 169-174. https://doi.org/10.11896/jsjkx.200500096

XU Hua-jie, ZHANG Chen-qiang, SU Guo-shao. Accurate Segmentation Method of Aerial Photography Buildings Based on Deep Convolutional Residual Network[J]. Computer Science, 2021, 48(8): 169-174. https://doi.org/10.11896/jsjkx.200500096

参考文献

[1]CORNELIS N,BASTIAN L,KURT C,et al.3D Urban Scene Modeling Integrating Recognition and[J].International Journal of Computer Vision,2008,78(2/3):121-141.
[2]SAHOO P K,SOLTANI S,WONG A K C.A survey of thre-sholding techniques[J].Computer Vision Graphics & Image Processing,1988,41(2):233-260.
[3]REUTER M,BIASOTTI S,GIORGI D,et al.Discrete Laplace-Beltrami operators for shape analysis and segmentation[J].Computers & Graphics,2009,33(3):381-390.
[4]ROTHER C,KOLMOGOROV V,BLAKE A.Grabcut:Interactive foreground extraction using iterated graph cuts[J].ACM Transactions on Graphics (TOG),ACM,2004,23(3):309-314.
[5]PAPADOPOULOS D P,UIJLINGS J R R,KELLER F,et al.Extreme clicking for efficient object annotation[C]//IEEE International Conference on Computer Vision.2017:4930-4939.
[6]MANINIS K K,CAELLES S,PONT-TUSET J,et al.Deep extreme cut:From extreme points to object segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:616-625.
[7]KRIZHEVSKY A,SUTSKEVER L,HINTON G E.ImageNetclassification with deep convolutional neural networks[C]//International Conference on Neural Information Processing Systems.Curran Associates Inc.2012:1097-1105.
[8]PAPANDREOU G,ZHU T,KANAZAWA N,et al.Towardsaccurate multi-person pose estimation in the wild[C]//Procee-dings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:4903-4911.
[9]KHORBOTLY S,HASSAN F.A modified approximation of 2D Gaussian smoothing filters for fixed-point platforms[C]//the IEEE 43rd Southeastern Symposium on System Theory.IEEE,2011:151-159.
[10]HE K,ZHANG X,REN S,et al.Deep residual learning forimage recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:770-778.
[11]CHEN L C,PAPANDREOU G,KOKKINOS I,et al.DeepLab:Semantic Image Segmentation with Deep Convolutional Nets,Atrous Convolution,and Fully Connected CRFs[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2018,40(4):834-848.
[12]ZHAO Z C,LUO Z,WANG P Y,et al.Survey on Image Classi-fication Algorithms Based on Deep Residual Network[J].Computer Systems & Applications,2020,29(1):14-21.
[13]YU F,KOLTUN V,FUNKHOUSER T.Dilated residual net-works[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:472-480.
[14]ZHAO H,SHI J,QI X,et al.Pyramid scene parsing network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:2881-2890.
[15]TAN C,SUN F,KONG T,et al.A survey on deep transfer learning[C]//International Conference on Artificial Neural Networks.Cham:Springer,2018:270-279.

相关文章 15

[1]	祝一帆, 王海涛, 李可, 吴贺俊. 一种高精度路面裂缝检测网络结构:Crack U-Net Crack U-Net:Towards High Quality Pavement Crack Detection 计算机科学, 2022, 49(1): 204-211. https://doi.org/10.11896/jsjkx.210100128
[2]	叶中玉, 吴梦麟. 融合时序监督和注意力机制的脉络膜新生血管分割 Choroidal Neovascularization Segmentation Combining Temporal Supervision and Attention Mechanism 计算机科学, 2021, 48(8): 118-124. https://doi.org/10.11896/jsjkx.200600150
[3]	胡育诚, 芮挺, 杨成松, 王东, 刘恂. 基于改进SIFT的无人机航拍图像快速配准研究 Study on Aerial Image Fast Registration from UAV 计算机科学, 2021, 48(8): 134-138. https://doi.org/10.11896/jsjkx.200600140
[4]	金海燕, 彭晶, 周挺, 肖照林. 基于Graph Cuts多特征选择的双目图像分割方法 Binocular Image Segmentation Based on Graph Cuts Multi-feature Selection 计算机科学, 2021, 48(8): 150-156. https://doi.org/10.11896/jsjkx.200800221
[5]	杨秀璋, 武帅, 夏换, 于小民. 基于自适应图像增强技术的水族文字提取与识别研究 Research on Shui Characters Extraction and Recognition Based on Adaptive Image Enhancement Technology 计算机科学, 2021, 48(6A): 74-79. https://doi.org/10.11896/jsjkx.200900070
[6]	曹林, 于威威. 基于图像分割的自适应窗口双目立体匹配算法研究 Adaptive Window Binocular Stereo Matching Algorithm Based on Image Segmentation 计算机科学, 2021, 48(11A): 314-318. https://doi.org/10.11896/jsjkx.201200264
[7]	顾兴健, 朱剑峰, 任守纲, 熊迎军, 徐焕良. 多尺度U网络实现番茄叶部病斑分割与识别 Multi-scale U Network Realizes Segmentation and Recognition of Tomato Leaf Disease 计算机科学, 2021, 48(11A): 360-366. https://doi.org/10.11896/jsjkx.201000166
[8]	杨志伟, 戴铭, 周智恒. 基于直方图差异的工业产品表面缺陷检测方法 Surface Defect Detection Method of Industrial Products Based on Histogram Difference 计算机科学, 2020, 47(6A): 247-249. https://doi.org/10.11896/JsJkx.191000049
[9]	杨连平, 孙玉波, 张红良, 李封, 张祥德. 基于编解码残差的人体关键点匹配网络 Human Keypoint Matching Network Based on Encoding and Decoding Residuals 计算机科学, 2020, 47(6): 114-120. https://doi.org/10.11896/jsjkx.200300079
[10]	曹义亲, 段也钰, 武丹. 基于WFSOA的2D-Otsu钢轨缺陷图像分割方法 2D-Otsu Rail Defect Image Segmentation Method Based on WFSOA 计算机科学, 2020, 47(5): 154-160. https://doi.org/10.11896/jsjkx.190200295
[11]	饶梦,苗夺谦,罗晟. 一种粗糙不确定的图像分割方法 Rough Uncertain Image Segmentation Method 计算机科学, 2020, 47(2): 72-75. https://doi.org/10.11896/jsjkx.190500177
[12]	雷涛,连倩,加小红,刘鹏. 基于快速SLIC的图像超像素算法 Fast Simple Linear Iterative Clustering for Image Superpixel Algorithm 计算机科学, 2020, 47(2): 143-149. https://doi.org/10.11896/jsjkx.190400121
[13]	周岳勇,程江华,刘通,王洋,陈明辉. 高分辨率SAR图像道路提取综述 Review of Road Extraction for High-resolution SAR Images 计算机科学, 2020, 47(1): 124-135. https://doi.org/10.11896/jsjkx.190100033
[14]	王嫣然, 陈清亮, 吴俊君. 面向复杂环境的图像语义分割方法综述 Research on Image Semantic Segmentation for Complex Environments 计算机科学, 2019, 46(9): 36-46. https://doi.org/10.11896/j.issn.1002-137X.2019.09.005
[15]	刘长齐, 邵堃, 霍星, 范冬阳, 檀结庆. 基于加权质量评价函数的K-means图像分割算法 K-means Image Segmentation Algorithm Based on Weighted Quality Evaluation Function 计算机科学, 2019, 46(6A): 158-160.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于深层卷积残差网络的航拍图建筑物精确分割方法

Accurate Segmentation Method of Aerial Photography Buildings Based on Deep Convolutional Residual Network

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0