A pre-trained model-based method for discriminating morphological patterns of vector-based coastlines

doi:10.11947/j.AGCS.2026.20250352

Abstract

Abstract:

Discriminating the morphological patterns of vector-based coastlines is vital for monitoring coastal evolution, marine disaster forecasting, and coastal zone planning, and it also serves as an important step in coastline cartography. Discriminating methods based on traditional machine learning technique rely on manually defined features and require large amounts of labeled samples with long-term training. To overcome these drawbacks, this study proposes a pre-trained model-based method that decouples generic geometric feature learning of coastlines from downstream morphological pattern discrimination. First, the coastlines are represented as Token sequences suitable for embedding learning using the operations of coordinate system resetting and coordinate normalization. Then, a self-supervised coordinate prediction task based on random masking is designed and integrated into the BERT model to construct a pre-trained model for the embedding learning of coastline geometric features. Finally, the pre-trained BERT model is fine-tuned with labeled dataset and transferred to the morphological pattern discrimination task. Based on open-source coastline data, a pre-trained dataset containing 195 649 samples and a labeled dataset with 1000 samples were collected. The proposed method achieves anF₁ score of 90.72% in a discrimination task involving five types of coastline morphological patterns, outperforming methods based on LSTM and 1D-CNN by 7.31%~9.38%.

Key words: vector-based coastline, morphological pattern discrimination, pre-trained model, BERT model

CLC Number:

P208

Min YANG, Hongran MA, Bo KONG, Pengcheng LIU, Tinghua AI. A pre-trained model-based method for discriminating morphological patterns of vector-based coastlines[J]. Acta Geodaetica et Cartographica Sinica, 2026, 55(3): 404-414.

Figures/Tables 12

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Tab. 1

Fig. 9

Tab. 2

Tab. 3

References 33

[1]	骆永明. 中国海岸带可持续发展中的生态环境问题与海岸科学发展[J]. 中国科学院院刊, 2016, 31(10): 1133-1142.
	LUO Yongming. Sustainability associated coastal eco-environmental problems and coastal science development in China[J]. Bulletin of the Chinese Academy of Sciences, 2016, 31(10): 1133-1142.
[2]	JOHNSON D W. Shore processes and shoreline development[M]. New York: Wiley, 1919: 159-190.
[3]	修淳, 霍素霞, 姚海燕, 等. 注重新形势下自然岸线管控的海岸线分类体系探讨[J]. 地理科学, 2022, 42(2): 333-342.
	XIU Chun, HUO Suxia, YAO Haiyan, et al. Coastline classification for strictly controlling natural coastline in the new era[J]. Scientia Geographica Sinica, 2022, 42(2): 333-342.
[4]	贾建军, 蔡廷禄, 刘毅飞, 等. 考虑人类活动的海岸线分类体系：近期浙江省海岸线调查的实践与思考[J]. 海洋科学, 2019, 43(10): 13-23.
	JIA Jianjun, CAI Tinglu, LIU Yifei, et al. A classification of coastline considering the impacts of human activities: remarks on latest practices on coastline survey in Zhejiang province[J]. Marine Sciences, 2019, 43(10): 13-23.
[5]	AI Tinghua, ZHOU Qi, ZHANG Xiang, et al. A simplification of ria coastline with geomorphologic characteristics preserved[J]. Marine Geodesy, 2014, 37(2): 167-186.
[6]	杜佳威, 武芳, 李靖涵, 等. 一种河口湾海岸线渐进化简方法[J]. 测绘学报, 2018, 47(4): 547-556. DOI: . doi: 10.11947/j.AGCS.2018.20170440
	DU Jiawei, WU Fang, LI Jinghan, et al. A progressive simplification method for the estuary coastline[J]. Acta Geodaetica et Cartographica Sinica, 2018, 47(4): 547-556. DOI: . doi: 10.11947/j.AGCS.2018.20170440
[7]	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.
[8]	MANDELBROT B. How long is the coast of Britain statistical self-similarity and fractional dimension[J]. Science, 1967, 156(3775): 636-638.
[9]	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. DOI: . doi: 10.11947/j.AGCS.2021.20210091
	AI Tinghua. Some thoughts on deep learning enabling cartography[J]. Acta Geodaetica et Cartograpgica Sinica, 2021, 50(9): 1170-1182. DOI: . doi: 10.11947/j.AGCS.2021.20210091
[10]	刘鹏程, 黄欣, 马宏然, 等. 建筑物多边形高精度识别的傅里叶形状描述子神经网络方法[J]. 测绘学报, 2022, 51(9): 1969-1976. DOI: . doi: 10.11947/j.AGCS.2022.20210730
	LIU Pengcheng, HUANG Xin, MA Hongran, et al. Fourier descriptor-based neural network method for high-precision shape recognition of building polygon[J]. Acta Geodaetica et Cartograpgica Sinica, 2022, 51(9): 1969-1976. DOI: . doi: 10.11947/j.AGCS.2022.20210730
[11]	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. DOI: . doi: 10.11947/j.AGCS.2022.20210302
	YAN Xiongfeng, YUAN Tuo, YANG Min, et al. An adaptive building simplification approach based on shape analysis and representation[J]. Acta Geodaetica et Cartograpgica Sinica, 2022, 51(2): 269-278. DOI: . doi: 10.11947/j.AGCS.2022.20210302
[12]	于洋洋, 贺康杰, 武芳, 等. 面状居民地形状分类的图卷积神经网络方法[J]. 测绘学报, 2022, 51(11): 2390-2402. DOI: . doi: 10.11947/j.AGCS.2022.20210134
	YU Yangyang, HE Kangjie, WU Fang, et al. Graph convolution neural network method for shape classification of areal settlements[J]. Acta Geodaetica et Cartograpgica Sinica, 2022, 51(11): 2390-2402. DOI: . doi: 10.11947/j.AGCS.2022.20210134
[13]	张付兵, 孙群, 马京振, 等. 融合全局和局部特征的建筑物形状智能分类方法[J]. 测绘学报, 2024, 53(9): 1842-1852. DOI: . doi: 10.11947/j.AGCS.2024.20240040
	ZHANG Fubing, SUN Qun, MA Jingzheng, et al. An intelligent classification method for building shape based on fusion of global and local features[J]. Acta Geodaetica et Cartograpgica Sinica, 2024, 53(9): 1842-1852. DOI: . doi: 10.11947/j.AGCS.2024.20240040
[14]	YANG Min, JIANG Chenjun, YAN Xiongfeng, et al. Detecting interchanges in road networks using a graph convolutional network approach[J]. International Journal of Geographical Information Science, 2022, 36(6): 1119-1139.
[15]	何海威, 钱海忠, 谢丽敏, 等. 立交桥识别的CNN卷积神经网络法[J]. 测绘学报, 2018, 47(3): 385-395. DOI: . doi: 10.11947/j.AGCS.2018.20170265
	HE Haiwei, QIAN Haizhong, XIE Limin, et al. Interchange recognition method based on CNN[J]. Acta Geodaetica et Cartograpgica Sinica, 2018, 47(3): 385-395. DOI: . doi: 10.11947/j.AGCS.2018.20170265
[16]	YU Huafei, AI Tinghua, Yang Min, et al. A recognition method for drainage patterns using a graph convolutional network[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 107: 102696.
[17]	余华飞, 邱天奇, 周哲, 等. 图Transformer支持下的河网模式识别[J]. 测绘学报, 2024, 53(11): 2075-2085. DOI: . doi: 10.11947/j.AGCS.2024.20230582
	YU Huafei, QIU Tianqi, ZHOU Zhe, et al. Drainage pattern recognition supported by graph Transformer[J]. Acta Geodaetica et Cartograpgica Sinica, 2024, 53(11): 2075-2085. DOI: . doi: 10.11947/j.AGCS.2024.20230582
[18]	孔博, 艾廷华, 杨敏, 等. 等高线形态知识与图神经网络联合作用下的黄土地貌类型识别[J]. 测绘学报, 2024, 53(6): 1154-1164. DOI: . doi: 10.11947/j.AGCS.2024.20230445
	KONG Bo, AI Tinghua, YANG Min, et al. Identification of loess landform types jointly affected by contour morphological knowledge and the graph neural network[J]. Acta Geodaeticaet Cartograpgica Sinica, 2024, 53(6): 1154-1164. DOI: . doi: 10.11947/j.AGCS.2024.20230445
[19]	GAO Aji, AI Tinghua, YU Huafei, et al. A vector-based coastline shape classification approach using sequential deep learning model[J]. International Journal of Applied Earth Observation and Geoinformation, 2024, 129: 103810.
[20]	杨敏, 陈果, 李连营, 等. 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. DOI: . doi: 10.11947/j.AGCS.2023.20210317
	YANG Min, CHEN Guo, LI Lianying, et al. Segmentation of linear map objects using sequential convolutional neural network[J]. Acta Geodaetica et Cartograpgica Sinica, 2023, 52(1): 108-116. DOI: . doi: 10.11947/j.AGCS.2023.20210317
[21]	PATWARDHAN N, MARRONE S, SANSONE C. Transformers in the real world: a survey on NLP applications[J]. Information, 2023, 14(4): 242.
[22]	SCHMIRLER R, HEINZINGER M, ROST B. Fine-tuning protein language models boosts predictions across diverse tasks[J]. Nature Communications, 2024, 15(1): 7407.
[23]	DEVLIN J, CHANG M-W, LEE K, et al. Bert: pre-training of deep bidirectional transformers for language understanding[C]//Proceedings of 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Minnesota: Association for Computational Linguistics, 2019: 4171-4186.
[24]	YAN Xiongfeng, YANG Min, AI Tinghua. Deep learning in automatic map generalization: achievements and challenges[J]. Geo-spatial Information Science, 2025, 28(6): 2905-2926.
[25]	MIKOLOV T, SUTSKEVER I, CHEN K, et al. Distributed representations of words and phrases and their compositionality[C]//Proceedings of the 27th Annual Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2013: 3111-3119.
[26]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2017: 6000-6010.
[27]	LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[28]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[29]	VAN DER MAATEN L. Accelerating t-SNE using tree-based algorithms[J]. The Journal of Machine Learning Research, 2014, 15(1): 3221-3245.
[30]	HUBERT L, ARABIE P. Comparing partitions[J]. Journal of Classification, 1985, 2(1): 193-218.
[31]	VINH N X, EPPS J, BAILEY J. Information theoretic measures for clusterings comparison: variants, properties, normalization and correction for chance[J]. Journal of Machine Learning Research, 2010(11): 2837-2854.
[32]	STEINLEY D. Properties of the hubert-arabie adjusted rand index[J]. Psychological Methods, 2004, 9(3): 386-396.
[33]	刘鹏程, 艾廷华, 杨敏. 基于傅里叶级数的等高线网络渐进式传输模型[J]. 测绘学报, 2012, 41(2): 284-290.
	LIU Pengcheng, AI Tinghua, YANG Min. The internet progressive transmission model for contour based on Fourier series[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(2): 284-290.

方法	精准率	召回率	F₁值
1D-CNN	81.76	81.50	81.34
LSTM	83.81	83.25	83.41
BERT	90.83	90.75	90.72

案例	海岸线形态	方法	判别概率
案例	海岸线形态	方法	光滑型	粗糙型	狭长型	宽谷型	人工型
案例1（狭长型）		1D-CNN	0.00	44.07	55.93	0.00	0.00
		LSTM	2.19	63.31	34.38	0.00	0.12
		BERT	0.00	13.24	86.76	0.00	0.00
案例2（宽谷型）		1D-CNN	0.00	57.04	0.22	41.05	1.69
		LSTM	2.48	33.19	1.30	61.05	1.98
		BERT	0.00	17.00	0.00	79.13	3.87
案例3（人工型）		1D-CNN	0.00	0.00	10.27	56.09	33.64
		LSTM	0.00	2.20	0.30	22.14	75.36
		BERT	0.00	1.66	0.17	0.00	98.17

案例	海岸线形态	方法	判别概率
案例	海岸线形态	方法	光滑型	粗糙型	狭长型	宽谷型	人工型
案例1（狭长型）		1D-CNN	0.00	9.46	15.72	68.31	6.51
		LSTM	0.00	7.22	28.25	64.53	0.00
		BERT	0.00	16.34	35.92	47.74	0.00
案例2（人工型）		1D-CNN	3.21	73.13	0.25	10.40	13.01
		LSTM	0.79	63.56	2.06	20.93	12.66
		BERT	0.00	84.04	0.00	0.00	15.96