干旱区生态环境知识资源中心(Arid): Vegetation type and land cover mapping in a semi-arid heterogeneous forested wetland of India: comparing image classification algorithms

Arid

DOI	10.1007/s10668-021-01596-6
	Vegetation type and land cover mapping in a semi-arid heterogeneous forested wetland of India: comparing image classification algorithms
	Deval, Kundan; Joshi, P. K.
通讯作者	Joshi, PK (corresponding author), Jawaharlal Nehru Univ, Sch Environm Sci, New Delhi 110067, India. ; Joshi, PK (corresponding author), Jawaharlal Nehru Univ, Special Ctr Disaster Res, New Delhi 110067, India.
来源期刊	ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY
ISSN	1387-585X
EISSN	1573-2975
出版年	2021-06
英文摘要	The present study evaluates and compares performance of three supervised classification algorithms namely Maximum Likelihood (MXL), Artificial Neural Network (ANN) and Support Vector Machine (SVM), using very high resolution WorldView-2 satellite imagery for vegetation type/land cover (VT/LC) mapping in Keoladeo National Park (KNP), India. We mapped 16 (8 gregarious VT and 8 LC) classes, and used Bootstrap (with 100 iterations) method for accuracy assessment. All three algorithms produced high overall accuracy (OA) (67-85%) and kappa (K) (65-83) values. Visual comparison of the predictions revealed that SVM (OA = 85.12% (K = 83.9) with 3.85% width of confidence interval) performed the best followed by ANN (69.72% (67.32) with 4.43%) and MXL (67.37% (65.22) with 4.33%). This research provides insight for selection of classification algorithm for detailed VT/LC mapping of wetland associated systems using very high resolution satellite data. The findings of this research are useful for environmental management, restoration and conservation planning of KNP, India. The database will be of high value for future development and sustainability issues in the park.
英文关键词	Vegetation type Land cover Semi-arid wetland Maximum Likelihood (MXL) Artificial Neural Network (ANN) Support Vector Machine (SVM) WorldView2
类型	Article ; Early Access
语种	英语
收录类别	SCI-E
WOS记录号	WOS:000664808200003
WOS关键词	SUPPORT VECTOR MACHINES ; ARTIFICIAL NEURAL-NETWORK ; KEOLADEO NATIONAL-PARK ; WORLDVIEW-2 ; ACCURACY ; CLASSIFIERS ; AREAS ; TREES
WOS类目	Green & Sustainable Science & Technology ; Environmental Sciences
WOS研究方向	Science & Technology - Other Topics ; Environmental Sciences & Ecology
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/367529
作者单位	[Deval, Kundan; Joshi, P. K.] Jawaharlal Nehru Univ, Sch Environm Sci, New Delhi 110067, India; [Joshi, P. K.] Jawaharlal Nehru Univ, Special Ctr Disaster Res, New Delhi 110067, India
推荐引用方式 GB/T 7714	Deval, Kundan,Joshi, P. K.. Vegetation type and land cover mapping in a semi-arid heterogeneous forested wetland of India: comparing image classification algorithms[J],2021.
APA	Deval, Kundan,&Joshi, P. K..(2021).Vegetation type and land cover mapping in a semi-arid heterogeneous forested wetland of India: comparing image classification algorithms.ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY.
MLA	Deval, Kundan,et al."Vegetation type and land cover mapping in a semi-arid heterogeneous forested wetland of India: comparing image classification algorithms".ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY (2021).