序列卷积神经网络支持下线状地图目标的分段方法

doi:10.11947/j.AGCS.2023.20210317

测绘学报 ›› 2023, Vol. 52 ›› Issue (1): 108-116.doi: 10.11947/j.AGCS.2023.20210317

序列卷积神经网络支持下线状地图目标的分段方法

杨敏¹, 陈果¹, 李连营¹, 黄浩然¹, 苗静³, 晏雄锋^1,2

1. 武汉大学资源与环境科学学院, 湖北武汉 430079;
2. 同济大学测绘与地理信息学院, 上海 200092;
3. 武汉市测绘研究院, 湖北武汉 430022

收稿日期:2021-06-07 修回日期:2022-05-18 发布日期:2023-02-09
通讯作者: 晏雄锋 E-mail:xiongfengyan@tongji.edu.cn
作者简介:杨敏(1985—),男,博士,副教授,研究方向为地图综合、深度学习。E-mail: yangmin2003@whu.edu.cn
基金资助:
国家自然科学基金(42071450;42001415);自然资源部数字制图与国土信息应用重点实验室开放研究基金(ZRZYBWD202101)

Segmentation of linear map objects using sequential convolutional neural network

YANG Min¹, CHEN Guo¹, LI Lianying¹, HUANG Haoran¹, MIAO Jing³, YAN Xiongfeng^1,2

1. School of Resource and Environmental Sciences, Wuhan University, Wuhan 430079, China;
2. College of Surveying and Geo-Informatics, Tongji University, Shanghai 200092, China;
3. Wuhan Geomatics Institute, Wuhan 430022, China

Received:2021-06-07 Revised:2022-05-18 Published:2023-02-09
Supported by:
The National Natural Science Foundation of China (Nos. 42071450;42001415);The Open Research Fund Program of Key Laboratory of Digital Mapping and Land Information Application Engineering, Ministry of Natural Resources (No.ZRZYBWD202101)

摘要/Abstract

摘要： 依据形态特征差异实施分段处理是实现线状地图目标自适应综合的重要步骤。传统方法主要采用长度、角度、曲率等参量描述局部曲线段的几何形态变化,通过人工设置规则或常规机器学习方法建立模式判别模型。本文提出基于卷积神经网络的线状目标模式识别与分段方法。首先,以相邻坐标点构成的线元为基本单元,以线元端点横、纵坐标差为特征,将线状目标离散化为二维序列;然后,建立序列卷积神经网络进行线元特征序列学习与预测,实现线元层次的模式类型判别;最后,利用迭代融合方法将拓扑相邻且模式类型相同的线元合并,从而输出不同形态模式的分段结果。以1∶5万行政区界线和1∶25万山区道路数据开展试验,本文方法分段结果与人工分段结果的一致性比率分别达到91.25%和85.65%,相较传统方法有一定提升。同时,本文方法通过深度学习获取模式判别的深层次特征,能够有效避免人工选择特征带来的主观性影响,对不同尺度、不同类型的线状目标分段问题适应性更好。

关键词: 线状地图目标, 分段, 序列卷积神经网络, 深度学习

Abstract: Segmentation of linear objects based on their morphological characteristics is a pivotal pre-step for adaptive generalization. Existing studies mainly use hand-crafted features, such as length, angle, and curvature, to describe the local structures of linear objects, and further to identify different patterns based on manual-defined rules or machine learning methods. In this study, we propose a structural recognition and segmentation method for linear objects using deep learning. First, a linear unit (also known as lixel) composed of two adjacent points is considered as the processing unit, and each linear object is discretized into a two-dimension sequence in which the differences between the horizontal and vertical coordinates of each lixel are encoded. Then, a sequential convolutional neural network (SCNN) is established to predict the types of each lixel. Finally, the segmentation results of different morphological characteristics are obtained by merging the adjacent lixels with the same type using an iteration method. Experiments were conducted on two datasets of 1∶50 k administrative boundaries and 1∶250 k mountain roads, and the consistency ratios of segmentation results were 91.25% and 85.65%, respectively, outperforming the traditional methods based on backpropagation artificial neural network and Naïve Bayes. Overall, our method can effectively avoid the subjectivity that exists when designing the hand-crafted features, and is more adaptable to the segmentation of linear objects with different scales and types.

Key words: linear map objects, segmentation, sequential convolutional neural network, deep learning

中图分类号:

P208

杨敏, 陈果, 李连营, 黄浩然, 苗静, 晏雄锋. 序列卷积神经网络支持下线状地图目标的分段方法[J]. 测绘学报, 2023, 52(1): 108-116.

YANG Min, CHEN Guo, LI Lianying, HUANG Haoran, MIAO Jing, YAN Xiongfeng. Segmentation of linear map objects using sequential convolutional neural network[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(1): 108-116.

参考文献

[1] BRASSEL K E, WEIBEL R. A review and conceptual framework of automated map generalization[J]. International Journal of Geographical Information Systems, 1988, 2(3): 229-244.
[2] SAMSONOV T E, YAKIMOVA O P. Shape-adaptive geometric simplification of heterogeneous line datasets[J]. International Journal of Geographical Information Science, 2017, 31(8): 1485-1520.
[3] AI Tinghua, KE Shu, YANG Min, et al. Envelope generation and simplification of polylines using Delaunay triangulation[J]. International Journal of Geographical Information Science, 2017, 31(2): 297-319.
[4] 许文帅, 龙毅, 周侗, 等. 基于邻近四点法的建筑物多边形化简[J]. 测绘学报, 2013, 42(6): 929-936. XU Wenshuai, LONG Yi, ZHOU Tong, et al. Simplification of building polygon based on adjacent four-point method[J]. Acta Geodaetica et Cartographica Sinica, 2013, 42(6): 929-936.
[5] YANG Min, YUAN Tuo, YAN Xiongfeng, et al. A hybrid approach to building simplification with an evaluator from a backpropagation neural network[J]. International Journal of Geographical Information Science, 2022, 36(2): 280-309.
[6] 尹烁, 闫小明, 晏雄锋. 基于特征边重构的建筑物化简方法[J]. 测绘学报, 2020, 49(6): 703-710. YIN Shuo, YAN Xiaoming, YAN Xiongfeng. Simplification method of building polygon based on feature edges reconstruction[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(6): 703-710.
[7] 晏雄锋, 袁拓, 杨敏, 等. 建筑物形状特征分析表达与自适应化简方法[J]. 测绘学报, 2022, 51(2): 269-278. YAN Xiongfeng, YUAN Tuo, YANG Min, et al. An adaptive building simplification approach based on shape analysis and representation[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(2): 269-278.
[8] 武芳, 巩现勇, 杜佳威. 地图制图综合回顾与前望[J]. 测绘学报, 2017, 46(10): 1645-1664. WU Fang, GONG Xianyong, DU Jiawei. Overview of the research progress in automated map generalization[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(10): 1645-1664.
[9] BALBOA J L G, LÓPEZ F J A. Generalization-oriented road line classification by means of an artificial neural network[J].GeoInformatica, 2008, 12(3): 289-312.
[10] 刘鹏程, 杨琴. 贝叶斯模型下面向地图表达的海岸线分段模型[J]. 计算机工程与应用, 2016, 52(22): 174-179. LIU Pengcheng, YANG Qin. Coastline segment model research for map generalization based on Bayesian method[J]. Computer Engineering and Applications, 2016, 52(22): 174-179.
[11] DOUGLAS D H, PEUCKER T K. Algorithms for the reduction of the number of points required to represent a digitized line or its caricature[J].Cartographica: the International Journal for Geographic Information and Geovisualization, 1973, 10(2): 112-122.
[12] VISVALINGAM M, WHYATT J D. Line generalisation by repeated elimination of points[J]. The Cartographic Journal, 1993, 30(1): 46-51.
[13] ARIZA LÓPEZ F J, GARCÍA BALBOA J L. Generalization-oriented road line segmentation by means of an artificial neural network applied over a moving window[J]. Pattern Recognition, 2008, 41(5): 1593-1609.
[14] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444.
[15] RONNEBERGER O, FISCHER P, BROX T. U-net:convolutional networks for biomedical image segmentation[M]. Cham: Springer International Publishing, 2015: 234-241.
[16] MINAEE S, BOYKOV Y, PORIKLI F, et al. Image segmentation using deep learning: a survey[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44(7): 3523-3542.
[17] KIM Y. Convolutional neural networks for sentence classification[C]//Proceedings of 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP). Doha, Qatar: Association for Computational Linguistics, 2014: 1746-1751.
[18] BRONSTEIN M M, BRUNA J, LECUN Y, et al. Geometric deep learning: going beyond euclidean data[J]. IEEE Signal Processing Magazine, 2017, 34(4): 18-42.
[19] YAN Xiongfeng, AI Tinghua, YANG Min, et al. Graph convolutional autoencoder model for the shape coding and cognition of buildings in maps[J]. International Journal of Geographical Information Science, 2021, 35(3): 490-512.
[20] DU Jiawei, WU Fang, XING Ruixing, et al. Segmentation and sampling method for complex polyline generalization based on a generative adversarial network[J]. Geocarto International, 2022, 37(14): 4158-4180.
[21] YAN Xiongfeng, AI Tinghua, YANG Min, et al. A graph convolutional neural network for classification of building patterns using spatial vector data[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2019, 150: 259-273.
[22] FENG Y, THIEMANN F, SESTER M. Learning cartographic building generalization with deep convolutional neural networks[J]. ISPRS International Journal of Geo-Information, 2019, 8(6): 258.
[23] YAN Xiongfeng, AI Tinghua, YANG Min, et al. A graph deep learning approach for urban building grouping[J]. Geocarto International, 2022, 37(10): 2944-2966.
[24] KINGMA D P, BA J. Adam: a method for stochastic optimization[C]//Proceedings of the 3rd International Conference for Learning Representations (ICLR). San Diego, California, USA:[s.n.], 2015.

序列卷积神经网络支持下线状地图目标的分段方法

Segmentation of linear map objects using sequential convolutional neural network

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	眭海刚, 刘畅, 干哲, 江政杰, 徐川. 多模态遥感图像匹配方法综述[J]. 测绘学报, 2022, 51(9): 1848-1861.
[2]	张良培, 何江, 杨倩倩, 肖屹, 袁强强. 数据驱动的多源遥感信息融合研究进展[J]. 测绘学报, 2022, 51(7): 1317-1337.
[3]	程涛, 张洋, James Haworth. 基于网络和图的时空智能——概念、方法和应用[J]. 测绘学报, 2022, 51(7): 1629-1639.
[4]	龚健雅, 宦麟茜, 郑先伟. 影像解译中的深度学习可解释性分析方法[J]. 测绘学报, 2022, 51(6): 873-884.
[5]	张勤, 赵超英, 陈雪蓉. 多源遥感地质灾害早期识别技术进展与发展趋势[J]. 测绘学报, 2022, 51(6): 885-896.
[6]	龚健雅, 张觅, 胡翔云, 张展, 李彦胜, 姜良存. 智能遥感深度学习框架与模型设计[J]. 测绘学报, 2022, 51(4): 475-487.
[7]	毛文婧, 王卫林, 焦利民, 刘安宝. 基于深度学习的中国连续空间覆盖PM_2.5浓度预报[J]. 测绘学报, 2022, 51(3): 361-372.
[8]	金飞, 官恺, 刘智, 韩佳容, 芮杰, 李庆高. 无监督密集匹配特征提取网络性能分析[J]. 测绘学报, 2022, 51(3): 426-436.
[9]	何直蒙, 丁海勇, 安炳琪. 高分辨率遥感影像建筑物提取的空洞卷积E-Unet算法[J]. 测绘学报, 2022, 51(3): 457-467.
[10]	吴琼, 葛大庆, 于峻川, 张玲, 李曼, 刘斌, 王艳, 马燕妮, 刘宏娟. 广域滑坡灾害隐患InSAR显著性形变区深度学习识别技术[J]. 测绘学报, 2022, 51(10): 2046-2055.
[11]	郭澳庆, 胡俊, 郑万基, 桂容, 杜志贵, 朱武, 贺乐和. 时序InSAR滑坡形变监测与预测的N-BEATS深度学习法——以新铺滑坡为例[J]. 测绘学报, 2022, 51(10): 2171-2182.
[12]	张永生, 张振超, 童晓冲, 纪松, 于英, 赖广陵. 地理空间智能研究进展和面临的若干挑战[J]. 测绘学报, 2021, 50(9): 1137-1146.
[13]	艾廷华. 深度学习赋能地图制图的若干思考[J]. 测绘学报, 2021, 50(9): 1170-1182.
[14]	龚健雅, 许越, 胡翔云, 姜良存, 张觅. 遥感影像智能解译样本库现状与研究[J]. 测绘学报, 2021, 50(8): 1013-1022.
[15]	杨必胜, 韩旭, 董震. 适用于城市场景大规模点云语义标识的深度学习网络[J]. 测绘学报, 2021, 50(8): 1059-1067.