A comparative analysis of different pixel and object-based classification algorithms using multi-source high spatial resolution satellite data for LULC mapping

被引:42
作者
Balha, Akanksha [1 ]
Mallick, Javed [2 ]
Pandey, Suneel [3 ]
Gupta, Sandeep [4 ]
Singh, Chander Kumar [1 ]
机构
[1] TERI Sch Adv Studies, Dept Energy & Environm, 10 Inst Area, New Delhi 110070, India
[2] King Khalid Univ, Dept Civil Engn, Abha, Saudi Arabia
[3] Energy & Resource Inst TERI, Environm & Waste Management Div, New Delhi, India
[4] Indira Gandhi Natl Open Univ, Reg Ctr, Jammu, Jammu & Kashmir, India
关键词
Radial basis function; Sigmoid; Kernel; Random forests; Naive Bayes; Mc Nemar's test; SUPPORT VECTOR MACHINES; REMOTE-SENSING DATA; LAND-COVER CLASSIFICATION; RANDOM FOREST; SVM; ACCURACY; TREES;
D O I
10.1007/s12145-021-00685-4
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
The preparation of accurate LULC is of great importance as it is used in various studies ranging from change detection to geospatial modelling. Literature offers studies comparing different classification algorithms/approaches to prepare LULC maps. However, still there is a lack of studies that can provide a comprehensive analysis on widely used classification algorithms. Hence, in the present study, nine different pixel- and object-based classification algorithms have been used to compare their relative effectiveness in generating remotely sensed LULC maps. The algorithms include maximum likelihood, neural network, support vector machine (linear, polynomial, RBF (radial basis function), sigmoid kernels), random forest (RF) and Naive Bayes for pixel-based classification and maximum likelihood algorithm for object-based classification (OBC) approach. Additionally, the study has analysed the impact of different types of satellite datasets (i.e., high resolution image and resolution merged images of same resolution) on relative effectiveness of the algorithms in classifying the satellite imageries accurately. High resolution (5 m) satellite image LISS 4 MX70, resolution merged satellite images (5 m) LISS 3 merged with LISS 4 mono and LISS 3 merged with IRS-1D are employed for classification. 27 LULC maps (9 classification algorithms * 3 images) are evaluated for comparing classification algorithms. The accuracy assessment of the images is carried out using confusion matrix and Mc Nemar's test. It has been observed that (1) the performance of all classification algorithms differs from each other and (2) resolution merged data impacts classification accuracy differently when compared to other satellite image of same spatial resolution. RF and OBC are identified as potential classifiers with majority of datasets. The results suggest that due to heterogeneity in urban land-use, it is difficult to achieve higher overall accuracy in classifying a large urban area using 5 m resolution satellite dataset. Moreover, visual examination of LULC should be considered for choosing better classification approach as pixel-based approach produces salt-pepper effect in LULC, whereas OBC produces visually smoothened LULC and identifies even smaller objects in urban landscape. The comparative evaluation of different image types reveal that RF performs better with all images, however, the performance of OBC was found to be improved with original high-resolution data.
引用
收藏
页码:2231 / 2247
页数:17
相关论文
共 68 条
[1]  
Agratiotis P, 2015, INT ARCH PHOTOGRAMM, V40-3, P1, DOI [10.5194/isprsarchives-XL-3-W2-1-2015, 10.5194/isprsarchives-XL-5-W5-1-2015]
[2]  
Agresti A., 2002, Categorical data analysis, V2nd, P165, DOI [DOI 10.1002/0471249688, 10.1002/0471249688]
[3]  
Anderson J.R., 1976, LAND USE LAND COVER, P28
[4]  
[Anonymous], 2011, RASCLASS SUPERVISED
[5]  
Baatz M., 2004, ECOGNITION PROFESSIO
[6]   Multi-resolution, object-oriented fuzzy analysis of remote sensing data for GIS-ready information [J].
Benz, UC ;
Hofmann, P ;
Willhauck, G ;
Lingenfelder, I ;
Heynen, M .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2004, 58 (3-4) :239-258
[7]   Random forests [J].
Breiman, L .
MACHINE LEARNING, 2001, 45 (01) :5-32
[8]   A new process for the segmentation of high resolution remote sensing imagery [J].
Chen, Z. ;
Zhao, Z. ;
Gong, P. ;
Zeng, B. .
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2006, 27 (22) :4991-5001
[9]   Classification of hyperspectral remote-sensing data with primal SVM for small-sized training dataset problem [J].
Chi, Mingmin ;
Feng, Rui ;
Bruzzone, Lorenzo .
ADVANCES IN SPACE RESEARCH, 2008, 41 (11) :1793-1799
[10]   Classification of the wildland-urban interface: A comparison of pixel- and object-based classifications using high-resolution aerial photography [J].
Cleve, Casey ;
Kelly, Maggi ;
Kearns, Faith R. ;
Morltz, Max .
COMPUTERS ENVIRONMENT AND URBAN SYSTEMS, 2008, 32 (04) :317-326