基于L1范数混合主动轮廓的河流SAR图像分割

doi:10.11947/j.AGCS.2024.20220690

测绘学报 ›› 2024, Vol. 53 ›› Issue (8): 1598-1609.doi: 10.11947/j.AGCS.2024.20220690

基于L₁范数混合主动轮廓的河流SAR图像分割

邢一波¹(), 韩斌¹, 鲍秉坤²()

^1.南京邮电大学通信与信息工程学院，江苏　南京　210003
^2.南京邮电大学计算机学院，江苏　南京　210023

收稿日期:2022-12-06 发布日期:2024-09-25
通讯作者: 鲍秉坤 E-mail:njupt_xyb@163.com;njupt_xyb@163.com;bingkunbao@njupt.edu.cn
作者简介:邢一波（1997—），男，硕士生，研究方向为遥感图像处理。E-mail：njupt_xyb@163.com
基金资助:
国家自然科学基金(62325206);江苏省自然科学基金(BK20220392);南京邮电大学引进人才自然科学研究启动基金(NY222004)

River SAR image segmentation using L₁ norm based hybrid active contours

Yibo XING¹(), Bin HAN¹, Bingkun BAO²()

^1.School of Communications and Information Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
^2.School of Computer Science, Nanjing University of Posts and Telecommunications, Nanjing 210023, China

Received:2022-12-06 Published:2024-09-25
Contact: Bingkun BAO E-mail:njupt_xyb@163.com;njupt_xyb@163.com;bingkunbao@njupt.edu.cn
About author:XING Yibo (1997—), male, postgraduate, majors in remote sensing image processing. E-mail: njupt_xyb@163.com
Supported by:
The National Natural Science Foundation of China(62325206);Natural Science Foundation of Jiangsu Province(BK20220392);Natural Science Research Start-up Foundation of Recruiting Talents of Nanjing University of Posts and Telecommunications(NY222004)

摘要/Abstract

摘要：

为解决现有主动轮廓模型难以准确分割河流SAR图像的问题，提出一种基于L₁范数的混合主动轮廓模型。首先，计算轮廓曲线内外区域像素灰度的中值作为区域拟合中心，以抑制SAR图像中干扰区域对其准确性的影响；然后，利用L₁范数构建新的能量约束项并在模型能量泛函中引入边缘指示函数，进一步提升模型的分割性能；最后，将基于L₁范数的中值和均值能量约束项结合起来并添加额外的区域拟合中心约束项，以提高模型的整体稳定性。针对实际河流SAR图像进行分割试验，结果表明，与现有分割方法相比，本文模型能更准确、稳定地分割河流SAR图像。

关键词: 河流分割, SAR图像, 主动轮廓模型, 混合能量项, L₁范数

Abstract:

To solve the problem that the existing active contour models are difficult to segment river SAR images accurately, this paper presents a hybrid active contour model based on the L₁ norm. First, the median values of the pixel intensities in the inner and outer regions of the contour curve are calculated as the region fitting centers to suppress the influence of the interference regions in SAR images on their accuracies. Second, the L₁ norm is used to construct a new energy constraint term and the edge indicator function is introduced into the model's energy functional to further enhance the segmentation performance. Finally, the median and mean energy constraint terms based on the L₁ norm are combined and additional region-fitting center constraint terms are added to improve the overall stability of the model. The segmentation experiments on real river SAR images show that the proposed model can segment river SAR images more accurately and stably than the existing models.

Key words: river segmentation, SAR image, active contour model, hybrid energy term, L₁ norm

中图分类号:

P237

邢一波, 韩斌, 鲍秉坤. 基于L₁范数混合主动轮廓的河流SAR图像分割[J]. 测绘学报, 2024, 53(8): 1598-1609.

Yibo XING, Bin HAN, Bingkun BAO. River SAR image segmentation using L₁ norm based hybrid active contours[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(8): 1598-1609.

图/表 5

图1

表1

表2

图2

表3

参考文献 26

[1]	VIGNESH T, THYAGHARAJAN K K. Water bodies identification from multispectral images using Gabor filter, FCM and canny edge detection methods[C]//Proceedings of 2017 International Conference on Information Communication and Embedded Systems (ICICES). Chennai: IEEE, 2017.
[2]	李占利, 刘宇琦, 孙瑜, 等. 基于卡尔曼滤波的SAR图像边缘检测方法[J]. 图学学报, 2019, 40(5): 823-828.
	LI Zhanli, LIU Yuqi, SUN Yu, et al. Edge detection for SAR images based on Kalman filter[J]. Journal of Graphics, 2019, 40(5): 823-828.
[3]	SGHAIER M O, FOUCHER S, LEPAGE R. River extraction from high-resolution SAR images combining a structural feature set and mathematical morphology[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017, 10(3): 1025-1038.
[4]	胡炎, 单子力, 高峰. 基于增强指数加权均值比的SAR图像边缘检测算法[J]. 电子与信息学报, 2018, 40(5): 1166-1172.
	HU Yan, SHAN Zili, GAO Feng. Edge detection algorithm for SAR image based on enhanced ROEWA[J]. Journal of Electronics & Information Technology, 2018, 40(5): 1166-1172.
[5]	LUO Yuxiao, AN Daoxiang, WANG Wu, et al. Improved ROEWA SAR image edge detector based on curvilinear structures extraction[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 17(4): 631-635.
[6]	LIANG Jiayong, LIU Desheng. A local thresholding approach to flood water delineation using Sentinel-1 SAR imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 159: 53-62.
[7]	杨蕴, 李玉, 赵泉华. 斑点抑制与多分辨率拓扑分析相结合的SAR图像河流水体提取[J]. 测绘学报, 2022, 51(1): 145-158. DOI: 10.11947/j.AGCS.2022.20190395.
	YANG Yun, LI Yu, ZHAO Quanhua. River waterbody extraction from SAR images based on speckle reduction and multi-resolution topological analysis[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(1): 145-158. DOI: 10.11947/j.AGCS.2022.20190395.
[8]	WANG Wenguang, WANG Jun, ZHAO Hui, et al. River detection from SAR images[C]//Proceedings the 5th Asia-Pacific Conference on Synthetic Aperture Radar (APSAR). Singapore: IEEE, 2015: 680-683.
[9]	GENG Jie, WANG Hongyu, FAN Jianchao, et al. Change detection of SAR images based on supervised contractive autoencoders and fuzzy clustering[C]//Proceedings of 2017 International Workshop on Remote Sensing with Intelligent Processing (RSIP). Shanghai: IEEE, 2017.
[10]	SHANG Ronghua, CHEN Chen, WANG Guangguang, et al. A thumbnail-based hierarchical fuzzy clustering algorithm for SAR image segmentation[J]. Signal Processing, 2020, 171: 107518.
[11]	CASELLES V, KIMMEL R, SAPIRO G. Geodesic active contours[C]//Proceedings of 1995 IEEE International Conference on Computer Vision. Cambridge: IEEE, 1995: 694-699.
[12]	方莉娜, 卢丽靖, 赵志远, 等. 车载激光点云道路边界提取的Snake方法[J]. 测绘学报, 2020, 49(11): 1438-1450. DOI: 10.11947/j.AGCS.2020.20190370.
	FANG Lina, LU Lijing, ZHAO Zhiyuan, et al. Road boundaries extraction from mobile laser scanning point clouds based on discrete point Snake[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(11): 1438-1450. DOI: 10.11947/j.AGCS.2020.20190370.
[13]	LÜ Hongli, ZHANG Yilin, WANG Renfang. Active contour model based on local absolute difference energy and fractional-order penalty term[J]. Applied Mathematical Modelling, 2022, 107: 207-232.
[14]	CHAN T F, VESE L A. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2): 266-277.
[15]	LIU Huaxiang, FANG Jiangxiong, ZHANG Zijian, et al. A novel active contour model guided by global and local signed energy-based pressure force[J]. IEEE Access, 2020, 8: 59412-59426.
[16]	FANG Jiangxiong, LIU Huaxiang, LIU Jun, et al. Fuzzy region-based active contour driven by global and local fitting energy for image segmentation[J]. Applied Soft Computing, 2021, 100: 106982.
[17]	SONG Yu, WU Yiquan, DAI Yimian. A new active contour remote sensing river image segmentation algorithm inspired from the cross entropy[J]. Digital Signal Processing, 2016, 48: 322-332.
[18]	HAN Bin, WU Yiquan. A novel active contour model based on modified symmetric cross entropy for remote sensing river image segmentation[J]. Pattern Recognition, 2017, 67: 396-409.
[19]	韩斌, 吴一全. SAR图像河流提取的主动轮廓模型的稳健估计算法[J]. 测绘学报, 2020, 49(6): 777-786. DOI: 10.11947/j.AGCS.2020.20180423.
	HAN Bin, WU Yiquan. Robust estimation algorithm of active contour model for river extraction in SAR images[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(6): 777-786. DOI: 10.11947/j.AGCS.2020.20180423.
[20]	SHELHAMER E, LONG J, DARRELL T. Fully convolutional networks for semantic segmentation[C]//Proceedings of 2017 IEEE Transactions on Pattern Analysis and Machine Intelligence. [S.l.]: IEEE, 2017: 640-651.
[21]	XU Chuan, ZHANG Shanshan, ZHAO Bofei, et al. SAR image water extraction using the attention U-net and multi-scale level set method: flood monitoring in South China in 2020 as a test case[J]. Geo-spatial Information Science, 2022, 25(2): 155-168.
[22]	SHAMSOLMOALI P, ZAREAPOOR M, WANG Ruili, et al. A novel deep structure U-net for sea-land segmentation in remote sensing images[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, 12(9): 3219-3232.
[23]	HOLZMANN M, DAVARI A, SEEHAUS T, et al. Glacier calving front segmentation using attention U-net[C]//Proceedings of 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS. Brussels: IEEE, 2021: 3483-3486.
[24]	REN Yibin, LI Xiaofeng, YANG Xiaofeng, et al. Development of a dual-attention U-net model for sea ice and open water classification on SAR images[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 4010205.
[25]	OH J, MARTIN D R, HU Xiaoping. Partitioned edge-function-scaled region-based active contour (p-ESRAC): automated liver segmentation in multiphase contrast-enhanced MRI[J]. Medical Physics, 2014, 41(4): 041914.
[26]	LI Chunming, XU Chenyang, GUI Changfeng, et al. Distance regularized level set evolution and its application to image segmentation[J]. IEEE Transactions on Image Processing: a Publication of the IEEE Signal Processing Society, 2010, 19(12): 3243-3254.

模型	图像1		图像2		图像3		图像4		图像5		图像6		图像7		图像8		图像9		图像10
模型	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率
CV模型	52.48	75.81	83.56	49.18	78.97	28.74	65.77	91.92	68.42	80.27	72.11	68.60	47.53	91.25	46.78	80.79	61.75	77.32	51.29	83.68
GLSPF模型	62.27	62.98	64.87	67.99	68.55	41.09	58.26	94.24	54.05	87.77	61.42	75.98	55.43	90.14	56.56	77.40	66.78	74.70	58.76	81.31
FGLFE模型	74.68	55.69	75.56	62.34	78.32	28.68	64.21	92.26	63.97	76.95	74.56	65.36	51.89	90.61	50.29	79.72	60.95	77.61	54.42	82.92
CE模型	84.48	26.22	89.15	38.16	83.14	25.27	73.68	90.47	75.17	76.95	82.56	57.69	81.18	79.17	69.41	70.92	71.89	71.42	65.90	79.20
MSCE模型	85.34	24.68	91.49	30.78	86.61	21.99	82.50	87.17	77.84	73.91	86.72	50.53	84.75	75.52	87.97	47.89	72.29	70.60	74.94	72.70
WL₁L_P模型	98.34	2.31	94.30	19.42	98.81	1.13	93.42	70.02	88.89	63.35	94.01	28.02	93.69	52.13	94.78	25.83	90.91	44.12	90.11	53.45
LT模型	72.68	59.68	70.17	65.18	78.12	29.91	74.22	90.16	57.23	84.11	76.01	64.13	52.01	90.16	50.19	80.01	61.10	77.31	55.60	82.59
THFC模型	86.68	21.23	92.62	23.86	86.32	22.56	95.68	68.26	93.68	42.68	93.68	28.84	92.56	63.69	88.02	49.62	92.02	39.35	88.35	58.68
本文模型	98.55	1.04	98.52	2.12	97.17	2.58	97.83	11.93	98.31	6.37	98.65	2.41	98.54	6.97	97.52	4.22	97.25	5.79	98.02	6.22

模型	图像1	图像2	图像3	图像4	图像5	图像6	图像7	图像8	图像9	图像10
CV模型	1.97	1.23	0.74	1.83	0.81	1.21	2.56	1.81	2.32	2.96
GLSPF模型	5.58	3.86	3.02	4.82	3.21	4.76	7.65	5.32	6.87	9.16
FGLFE模型	4.06	2.45	2.08	3.76	2.06	4.22	5.32	4.01	4.87	6.03
CE模型	2.49	1.55	1.13	2.79	1.22	3.68	3.86	2.20	2.94	4.54
MSCE模型	2.89	1.64	1.34	3.25	1.43	4.03	4.12	2.89	3.54	5.13
WL₁L_P模型	1.54	1.17	1.02	1.18	0.75	1.15	2.20	1.27	2.39	1.92
LT模型	3.86	2.27	1.98	3.31	1.98	3.75	4.86	3.21	4.21	5.72
THFC模型	5.03	3.35	2.16	3.89	2.23	4.25	6.71	4.13	5.68	7.11
本文模型	1.67	1.16	1.01	1.13	0.68	1.02	2.13	1.07	2.01	1.98

图像	RDU－Net模型		AU－Net模型		DAU－Net模型		本文模型
图像	准确率	虚警率	准确率	虚警率	准确率	虚警率	准确率	虚警率
图像1	84.09	26.54	86.01	19.56	88.49	15.00	98.55	1.04
图像2	97.49	8.48	97.70	1.99	98.10	2.67	98.52	2.12
图像3	93.23	14.24	96.09	8.86	96.52	6.40	97.17	2.58
图像4	96.26	51.02	97.19	36.38	96.93	44.13	97.83	11.93
图像5	96.16	15.43	96.23	12.83	96.33	14.51	98.31	6.37
图像6	93.88	27.44	95.18	19.87	95.53	12.09	98.65	2.41
图像7	94.16	53.36	95.48	43.14	96.50	30.57	98.54	6.97
图像8	92.99	32.40	95.76	9.01	95.96	8.33	97.52	4.22
图像9	96.66	16.28	96.86	13.23	97.16	11.32	97.25	5.79
图像10	94.89	30.76	96.64	15.92	96.56	15.96	98.02	6.22
平均值	93.08	27.80	95.21	17.88	95.91	16.00	98.21	4.97

基于L₁范数混合主动轮廓的河流SAR图像分割

River SAR image segmentation using L₁ norm based hybrid active contours

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 26

相关文章 15

编辑推荐

Metrics

本文评价

[1]	项德良, 丁怀跃, 管冬冬, 程建达, 孙晓坤. 结合极化白化滤波和SimSD-CapsuleNet的PolSAR图像配准[J]. 测绘学报, 2024, 53(3): 450-462.
[2]	杨蕴, 李玉, 赵泉华. 斑点抑制与多分辨率拓扑分析相结合的SAR图像河流水体提取[J]. 测绘学报, 2022, 51(1): 145-158.
[3]	韩斌, 吴一全. SAR图像河流提取的主动轮廓模型的稳健估计算法[J]. 测绘学报, 2020, 49(6): 777-786.
[4]	李东宸, 向文豪, 党倩楠, 吴艳. 均匀分布与结构描述ASIFT相结合的SAR与可见光图像配准算法[J]. 测绘学报, 2020, 49(12): 1583-1590.
[5]	方莉娜, 卢丽靖, 赵志远, 王昀昀, 陈崇成. 车载激光点云道路边界提取的Snake方法[J]. 测绘学报, 2020, 49(11): 1438-1450.
[6]	闫广峰, 岑敏仪. L₁范数探测粗差失效的观测量识别方法[J]. 测绘学报, 2019, 48(11): 1430-1438.
[7]	王岱良, 李玉, 林文杰, 赵泉华. 任意方向对称差值核的SAR图像边缘提取算法[J]. 测绘学报, 2017, 46(9): 1165-1173.
[8]	韩斌, 吴一全. SAR图像河流分割的加权指数区域能量模型[J]. 测绘学报, 2017, 46(9): 1174-1181.
[9]	姜大伟, 范剑超, 黄凤荣. SAR图像海岸线检测的区域距离正则化几何主动轮廓模型[J]. 测绘学报, 2016, 45(9): 1096-1103.
[10]	陈继伟, 曾琪明, 焦健, 叶发旺, 朱黎江. 利用轨道参数修正的无控制点星载SAR图像几何校正方法[J]. 测绘学报, 2016, 45(12): 1434-1440.
[11]	苏娟张强陈炜王继平. 高分辨率SAR图像中建筑物特征融合检测算法[J]. 测绘学报, 2014, 43(9): 939-944.
[12]	孙颖张新长罗国玮. 从机载激光雷达点云提取建筑物屋顶边界的活动轮廓模型改进方法[J]. 测绘学报, 2014, 43(6): 620-636.
[13]	卜丽静张过. 联合稀疏约束的双通道星载SAR图像重建[J]. 测绘学报, 2014, 43(5): 521-528.
[14]	李新娜王正明谢美华. 一种基于核回归的SAR图像自适应相干斑抑制方法[J]. , 2013, 42(5): 729-737.
[15]	王文波,赵攀,张晓东. 利用经验模态分解和主成分分析的SAR图像相干斑抑制[J]. 测绘学报, 2012, 41(6): 838-843.