Object-Oriented Classification of Wetland Vegetation Community in Jilin-1 Remote Sensing Image

doi:10.13466/j.cnki.lyzygl.2023.01.017

Abstract

Abstract:

The use of remote sensing technology to extract the composition and distribution of wetland vegetation communities is of great significance to the construction of wetlands.Taking Qinghai Ulandulan Lake National Wetland Park as the research area,using the number of Jilin No.1 remote sensing images,KNN and RF classification models were selected through image segmentation and feature optimization,the vegetation community of Dulan Lake wetland was divided,and the classification accuracy was verified.The results showed that according to the segmentation scale provided by ESP 2 tool,the optimal segmentation scale for object-oriented classification of vegetation communities was 18,and the segmentation scales of vegetation and non-vegetation areas were 32 and 85,respectively.For character type division,only using the threshold of image band information and related index could not accurately extract the feature category,it was necessary to combine the image geometric features and texture features to improve the classification accuracy,use the feature space optimization tool to optimize 61 image features,and finally screen out 40 image features,and use them for classification.According to the confusion matrix classification accuracy evaluation results,the classification results of KNN algorithm were better than RF,among which the overall classification accuracy of KNN was 81.80%,the Kappa coefficient was 0.79,the overall classification accuracy of RF was 72.59%,and the Kappa coefficient was 0.68.According to the classification results,the vegetation coverage rate of Dulan Lake wetland was 44.41%,and the composition and distribution characteristics of vegetation communities in the results could provide a basis for wetland ecological construction and management.

Key words: Lake Dulan, vegetation community classification, object-oriented, image features, classification model

CLC Number:

S771.8

XIE Wenchun, LI Qiangfeng, LI Yanchun, WU Zhenshan, YANG Zhengfan. Object-Oriented Classification of Wetland Vegetation Community in Jilin-1 Remote Sensing Image[J]. FOREST RESOURCES WANAGEMENT, 2023, 0(1): 141-152.

Figures/Tables 11

Fig.1

Tab.1

Tab.2

Fig.2

Fig.3

Fig.4

Fig.5

Tab.3

Tab.4

Fig.6

Tab.5

References 36

[1]	吕宪国, 邹元春. 中国湿地研究[M]. 长沙: 湖南教育出版社, 2017.
[2]	董婵, 黄英豪, 闵凡路. 湿地植物与生态环境[M]. 天津: 天津科学技术出版社. 2014.
[3]	李旭, 谢永宏, 黄继山, 等. 湿地植被格局成因研究进展[J]. 湿地科学, 2009, 7(3):280-288.
[4]	郭柯, 方精云, 王国宏, 等. 中国植被分类系统修订方案[J]. 植物生态学报, 2020, 44(2):111-127. doi: 10.17521/cjpe.2019.0271
[5]	Roughgarden J, Running S W, Matson P A. What does remote sensing do for ecology?[J]. Ecology, 1991, 72(6):1918-1922. doi: 10.2307/1941546
[6]	Ozesmi S L, Bauer M E. Satellite remote sensing of wetlands[J]. Wetlands ecology and management, 2002, 10(5):381-402. doi: 10.1023/A:1020908432489
[7]	Schmidt K S, Skidmore A K, Kloosterman E H, et al. Mapping coastal vegetation using an expert system and hyperspectral imagery[J]. Photogrammetric Engineering & Remote Sensing, 2004, 70(6):703-715.
[8]	Szantoi Z, Escobedo F, Abd-Elrahman A, et al. Analyzing fine-scale wetland composition using high resolution imagery and texture features[J]. International Journal of Applied Earth Observation and Geoinformation, 2013, 23:204-212. doi: 10.1016/j.jag.2013.01.003
[9]	闫利, 江维薇. 多光谱遥感影像植被覆盖分类研究进展[J]. 国土资源遥感, 2016, 28(2):8-13.
[10]	张强. 湿地植被遥感分类研究进展[J]. 世界林业研究, 2019, 32(3):49-54.
[11]	谭衢霖, Steve Johansen. 基于像元和对象分类的城区植被高分辨率遥感制图比较研究[J]. 应用基础与工程科学学报, 2011, 19(3):441-448.
[12]	Johansen K, Coops N C, Gergel S E, et al. Application of high spatial resolution satellite imagery for riparian and forest ecosystem classification[J]. Remote Sensing of Environment, 2007, 110(1):29-44. doi: 10.1016/j.rse.2007.02.014
[13]	Gilmore M S, Wilson E H, Barrett N, et al. Integrating multi-temporal spectral and structural information to map wetland vegetation in a lower Connecticut River tidal marsh[J]. Remote Sensing of Environment, 2008, 112(11):4048-4060. doi: 10.1016/j.rse.2008.05.020
[14]	杨盼盼. 基于高分辨率遥感影像纹理特征的面向对象植被分类方法研究[D]. 昆明: 云南师范大学, 2017.
[15]	Carle M V, Wang L, Sasser C E. Mapping freshwater marsh species distributions using WorldView-2 high-resolution multispectral satellite imagery[J]. International journal of remote sensing, 2014, 35(13):4698-4716. doi: 10.1080/01431161.2014.919685
[16]	Bhatnagar S, Gill L, Regan S, et al. Mapping vegetation communities inside wetlands using Sentinel-2 imagery in Ireland[J/OL]. International Journal of Applied Earth Observation and Geoinformation, 2020, 88:102083.(2020-02-14)[2022-08-23]. https://www.sciencedirect.com/science/article/pii/S0303243419308359?via%3Dihub. doi: 10.1016/j.jag.2020.102083
[17]	Yeo S, Lafon V, Alard D, et al. Classification and mapping of saltmarsh vegetation combining multispectral images with field data[J/OL]. Estuarine,Coastal and Shelf Science, 2020, 236:106643.(2020-02-17)[2022-08-23]https://www.sciencedirect.com/science/article/abs/pii/S027277141930232X.
[18]	Ju Y, Bohrer G. Classification of wetland vegetation based on NDVI time series from the HLS dataset[J/OL]. Remote Sensing, 2022, 14(9):2107.(2022-04-27)[2022-08-25]https://www.mdpi.com/2072-4292/14/9/2107.
[19]	Lou Peiqing, Fu Bolin, He Hongchang, et al. An optimized object-based random forest algorithm for marsh vegetation mapping using high-spatial-resolution GF-1 and ZY-3 data[J/OL]. Remote Sensing, 2020, 12(8):1270.(2022-04-17)[2022-08-25]https://www.mdpi.com/2072-4292/12/8/1270.
[20]	谢锦莹, 丁丽霞, 王志辉, 等. 基于FCN与面向对象的滨海湿地植被分类[J]. 林业科学, 2020, 56(8):98-106.
[21]	刘尚武. 青海植物志[M]. 青海: 青海人民出版社,1996.
[22]	金伊丽, 王皓言, 魏临风, 等. 青藏高原植物群落样方数据集[J]. 植物生态学报, 2022, 46(7):846-854. doi: 10.17521/cjpe.2022.0174
[23]	GB/T 21010-2017,土地利用现状分类[S].
[24]	刘瑞清, 李加林, 孙超, 等. 基于Sentinel-2遥感时间序列植被物候特征的盐城滨海湿地植被分类[J]. 地理学报, 2021, 76(7):1680-1692. doi: 10.11821/dlxb202107008
[25]	邹文涛, 张怀清, 鞠洪波, 等. 基于QUEST树的高寒湿地植被覆盖类型遥感分类研究[J]. 中南林业科技大学学报, 2011, 31(12):138-144.
[26]	张学儒, 刘林山, 张镱锂, 等. 基于ENVI ZOOM面向对象的高海拔灌丛植被提取:以定日县为例[J]. 地理与地理信息科学, 2010, 26(4):104-108.
[27]	陈春雷, 武刚. 面向对象的遥感影像最优分割尺度评价[J]. 遥感技术与应用, 2011, 26(1):96-102.
[28]	郭亚鸽, 于信芳, 江东, 等. 面向对象的森林植被图像识别分类方法[J]. 地球信息科学学报, 2012, 14(4):514-522.
[29]	Benz U C, Hofmann P, Willhauck G, et al. Multi-resolution,object-oriented fuzzy analysis of remote sensing data for GIS-ready information[J]. ISPRS Journal of photogrammetry and remote sensing, 2004, 58(3-4):239-258. doi: 10.1016/j.isprsjprs.2003.10.002
[30]	邬亚娟, 刘廷玺, 童新, 等. 基于面向对象的干旱半干旱地区植被分类[J]. 干旱区研究, 2020, 37(4):1026-1034.
[31]	马浩然, 赵天忠, 曾怡. 面向对象的最优分割尺度下多层次森林植被分类[J]. 东北林业大学学报, 2014, 42(9):52-57.
[32]	陈丽萍, 孙玉军. 基于不同决策树的面向对象林区遥感影像分类比较[J]. 应用生态学报, 2018, 29(12):3995-4003. doi: 10.13287/j.1001-9332.201812.015
[33]	李哲, 张沁雨, 彭道黎. 基于高分二号遥感影像的树种分类方法[J]. 遥感技术与应用, 2019, 34(5):970-982.
[34]	宗影, 李玉凤, 刘红玉. 基于面向对象随机森林方法的滨海湿地植被分类研究[J]. 南京师范大学学报:工程技术版, 2021, 21(4):47-55.
[35]	张晨宇, 陈沈良, 李鹏, 等. 现行黄河口保护区典型湿地植被时空动态遥感监测[J]. 海洋学报, 2022, 44(1):125-136.
[36]	施佩荣, 陈永富, 刘华, 等. 基于改进的面向对象遥感影像分类方法研究:以西藏米林县典型林区为例[J]. 遥感技术与应用, 2017, 32(3):466-474.

植被类型	优势种	主要伴生种	采样点
沼泽草甸	芦苇(Phragmites australis)	盐角草(Salicornia europaea)	40
高寒灌丛	小果白刺(Nitraria sibirica)	大白刺(Nitraria roborowskii)	40
高寒灌丛	细枝盐爪爪(Kalidium gracile)+ 芨芨草(Achnatherum splendens)	猪毛菜Kali collinum)、驼绒藜(Krascheninnikovia ceratoides)、盐地碱蓬(Suaeda salsa)、红砂(Reaumuria songarica)	36
沼泽草甸	盐角草(Salicornia europaea)	盐爪爪(Kalidium foliatum)	30
高寒灌丛	枸杞(Lycium chinense)	小果白刺(Nitraria sibirica)	28
高寒草甸	早熟禾(Poa annua)+ 盐地风毛菊(Saussurea salsa)	海乳草(Lysimachia maritima)、海韭菜(Triglochin maritimum)、马蔺(Iris lactea)	28

特征类别	特征描述		数量
光谱特征	归一化植被指数(NDVI)	NDVI=(NIR-R)/(NIR+R)	16
	土壤修正植被指数(SAVI)	SAVI=(NIR-R)×(1+0.5)/(NIR+R+0.5)
	比值植被指数(RVI)	RVI=NIR/R
	归一化水体指数(NDWI)	NDWI=(G-NIR)/(G+NIR)
	土壤调节植被指数(OSAVI)	OSAVI=(NIR-R)/(NIR+R+0.16)
	增强型植被指数(EVI)	EVI=2.5×(NIR-R)/(NIR+6×R-7.5×B+1)
	亮度、波段均值、波段标准差、最大差异度量
纹理特征	均值、同质性、对比度、墒、相关性、差异性、标准差、角二阶矩		40
几何特征	面积、紧致度、长宽比、矩形匹配度、形状指数		5

序号	特征名称	序号	特征名称	序号	特征名称	序号	特征名称
1	Area	11	Mean G	21	GLCM Correlation(90°)	31	GLCM Homogeneity(45°)
2	GLCM Mean(0°)	12	Standard deviation R	22	GLCM Homogeneity(135°)	32	EVI
3	Brightness	13	Standard deviation NIR	23	Rectandular fit	33	OSAVI
4	GLCM Mean(45°)	14	Standard deviation G	24	Max.diff.	34	NDWI
5	Compactness	15	Standard deviation B	25	GLCM Entropy(90°)	35	RVI
6	GLCM Entropy(all dir.)	16	GLCM Homogeneity(90°)	26	GLCM Mean(90°)	36	SAVI
7	GLCM Mean(all dir.)	17	GLCM Entropy(45°)	27	GLCM Mean(135°)	37	NDVI
8	Shape index	18	GLCM Homogeneity(0°)	28	GLCM StdDev(135°)	38	GLCM Correlation(45°)
9	Mean NIR	19	GLCM Entropy(135°)	29	GLCM Correlation(135°)	39	GLCM Correlation(all dir.)
10	Mean B	20	Length/Width	30	GLCM Correlation(0°)	40	GLCM Entropy(0°)

类别	类别
类别	PA	BG	BW	NS	KG+AS	LC	IS	PA+SS	SE
PA	0	16.84	18.04	8.20	12.46	14.11	16.53	11.71	7.32
BG	16.84	0	13.18	7.59	7.26	7.56	3.31	4.05	5.33
BW	18.04	13.18	0	17.75	17.34	22.16	13.55	14.57	16.63
NS	8.20	7.59	17.75	0	2.34	2.13	6.55	2.44	2.29
KG+AS	12.46	7.26	17.34	2.34	0	3.03	4.40	3.05	4.06
LC	14.11	7.56	22.16	2.13	3.04	0	5.94	3.07	2.52
IS	16.53	3.31	13.55	6.55	4.40	5.94	0	4.30	5.83
PA+SS	11.71	4.05	14.57	2.44	3.05	3.07	4.30	0	1.74
SE	7.32	5.33	16.63	2.29	4.06	2.52	5.83	1.74	0

植被群落及精度				BG		BW		IS		KG+AS		LC		NS		PA		PA+SS		SE		UA/%
KNN		BG		32		2		1		0		0		3		0		4		0		76.19
		BW		3		51		0		0		0		0		1		0		0		92.73
		IS		3		0		7		0		0		0		0		0		0		70.00
		KG+AS		1		0		0		59		4		9		2		0		1		77.63
		LC		0		0		0		0		31		2		0		0		0		93.94
		NS		0		0		0		2		2		56		1		2		0		88.89
		PA		0		0		0		1		0		4		43		0		3		84.31
		PA+SS		0		0		0		5		0		3		2		64		9		77.11
		SE		0		0		0		2		0		1		1		9		30		69.77
		PA/%		82.05		96.23		87.50		85.51		83.78		71.79		86.00		81.01		69.77
				OA%=81.80%,Kappa系数=0.79
RF	BG		21		2		1		5		0		3		0		4		0		58.33
	BW		0		51		0		0		0		0		0		0		0		100.00
	IS		5		0		5		0		0		0		0		0		0		50.00
	KG+AS		8		0		0		44		0		13		2		6		5		56.41
	LC		0		0		0		1		34		1		0		0		0		94.44
	NS		3		0		1		9		1		48		5		8		0		64.00
	PA		0		0		0		3		0		4		42		0		0		85.71
	PA+SS		2		0		1		3		2		8		0		59		11		68.60
	SE		0		0		0		4		0		1		1		2		27		77.14
	PA/%		53.85		96.23		62.50		63.77		91.89		61.54		84.00		74.68		62.79
			OA%=72.59%,Kappa系数=0.68