Comparison of the Applicability of J-M Distance Feature Selection Methods for Coastal Wetland Classification

Accurate determination of the spatial distribution of coastal wetlands is crucial for the management and conservation of ecosystems. Feature selection methods based on the Jeffries-Matusita (J-M) method include J-M distance with simple average ranking (JMave), J-M distance based on weights and correlations (JMimproved), and heuristic J-M distance (JMmc). However, as the impacts of these methods on wetland classification are different, their applicability has rarely been investigated. Based on the Google Earth Engine (GEE) and random forest (RF) classifier, this is a comparative analysis of the applicability of the JMave, JMimproved, and JMmc methods. The results show that the three methods compress feature dimensions and retain all feature types as much as possible. JMmc exhibits the most significant compression from a value of 35 to 15 (57.14%), which is 37.14% and 40% more compressed than JMave and JMimproved, respectively. Moreover, they produce comparable classification results, with an overall classification accuracy of 90.20 ± 0.19% and a Kappa coefficient of 88.80 ± 0.22%. However, different methods had their own advantages for the classification of different land classes. Specifically, JMave has a better classification only in cropland, while JMmc is advantageous for recognizing water bodies, tidal flats, and aquaculture. While JMimproved failed to retain vegetation and mangrove features, it enables a better depiction of the mangroves, salt pans, and vegetation classes. Both JMave and JMimproved rearrange features based on J-M distance, while JMmc places more emphasis on feature selection. As a result, there can be significant differences in feature subsets among these three methods. Therefore, the comparative analysis of these three methods further elucidates the importance of J-M distance in feature selection, demonstrating the significant potential of J-M distance-based feature selection methods in wetland classification.

[1]  D. Hou,et al.  A coastal wetlands mapping approach of Yellow River Delta with a hierarchical classification and optimal feature selection framework , 2023, CATENA.

[2]  Gang Zhao,et al.  Improving Satellite Retrieval of Coastal Aquaculture Pond by Adding Water Quality Parameters , 2022, Remote. Sens..

[3]  Qigang Jiang,et al.  Object-Based Multigrained Cascade Forest Method for Wetland Classification Using Sentinel-2 and Radarsat-2 Imagery , 2022, Water.

[4]  Basabi Chakraborty,et al.  An effective feature subset selection approach based on Jeffries-Matusita distance for multiclass problems , 2021, J. Intell. Fuzzy Syst..

[5]  Ehsan Farahbakhsh,et al.  A review of machine learning in processing remote sensing data for mineral exploration , 2021, Remote Sensing of Environment.

[6]  Yonggen Zhang,et al.  Mapping Coastal Wetlands of the Bohai Rim at a Spatial Resolution of 10 m Using Multiple Open-Access Satellite Data and Terrain Indices , 2020, Remote. Sens..

[7]  Saeid Parsian,et al.  Google Earth Engine Cloud Computing Platform for Remote Sensing Big Data Applications: A Comprehensive Review , 2020, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[8]  A. Blanco,et al.  Development and application of a new mangrove vegetation index (MVI) for rapid and accurate mangrove mapping , 2020, ISPRS Journal of Photogrammetry and Remote Sensing.

[9]  Rung-Ching Chen,et al.  Selecting critical features for data classification based on machine learning methods , 2020, Journal of Big Data.

[10]  Zhiguang Tang,et al.  Extraction and analysis of spatiotemporal variation of rice planting area in Hunan Province based on MODIS time-series data , 2020 .

[11]  Shrutika S. Sawant,et al.  A survey of band selection techniques for hyperspectral image classification , 2020, Journal of Spectral Imaging.

[12]  Chao Wang,et al.  Monitoring Invasion Process of Spartina alterniflora by Seasonal Sentinel-2 Imagery and an Object-Based Random Forest Classification , 2020, Remote. Sens..

[13]  Zongming Wang,et al.  A New Vegetation Index to Detect Periodically Submerged Mangrove Forest Using Single-Tide Sentinel-2 Imagery , 2019, Remote. Sens..

[14]  Qiusheng Wu,et al.  Integrating LiDAR data and multi-temporal aerial imagery to map wetland inundation dynamics using Google Earth Engine. , 2019, Remote sensing of environment.

[15]  Ugur Avdan,et al.  Evaluating the utilization of the red edge and radar bands from sentinel sensors for wetland classification , 2019, CATENA.

[16]  Yi Ma,et al.  Deep learning classification of coastal wetland hyperspectral image combined spectra and texture features: A case study of Huanghe (Yellow) River Estuary wetland , 2019, Acta Oceanologica Sinica.

[17]  Benoit Montpetit,et al.  Wetland Classification with Multi-Angle/Temporal SAR Using Random Forests , 2019, Remote. Sens..

[18]  Sudarson Jena,et al.  Correlation based feature selection with clustering for high dimensional data , 2018, Journal of Electrical Systems and Information Technology.

[19]  Brian Brisco,et al.  Remote sensing for wetland classification: a comprehensive review , 2018 .

[20]  Hui Lin,et al.  Improving Selection of Spectral Variables for Vegetation Classification of East Dongting Lake, China, Using a Gaofen-1 Image , 2017, Remote. Sens..

[21]  Lizhe Wang,et al.  Assessing Different Feature Sets' Effects on Land Cover Classification in Complex Surface-Mined Landscapes by ZiYuan-3 Satellite Imagery , 2017, Remote. Sens..

[22]  Lori White,et al.  Moving to the RADARSAT Constellation Mission: Comparing Synthesized Compact Polarimetry and Dual Polarimetry Data with Fully Polarimetric RADARSAT-2 Data for Image Classification of Peatlands , 2017, Remote. Sens..

[23]  Jie Tian,et al.  Random Forest Classification of Wetland Landcovers from Multi-Sensor Data in the Arid Region of Xinjiang, China , 2016, Remote. Sens..

[24]  Alfonso Fernández-Manso,et al.  SENTINEL-2A red-edge spectral indices suitability for discriminating burn severity , 2016, Int. J. Appl. Earth Obs. Geoinformation.

[25]  Mariana Belgiu,et al.  Random forest in remote sensing: A review of applications and future directions , 2016 .

[26]  O. Mutanga,et al.  Evaluating the impact of red-edge band from Rapideye image for classifying insect defoliation levels , 2014 .

[27]  K. Sundar,et al.  Can wetlands maintained for human use also help conserve biodiversity? Landscape-scale patterns of bird use of wetlands in an agricultural landscape in north India , 2013 .

[28]  Badih Ghattas,et al.  A review of supervised machine learning algorithms and their applications to ecological data , 2012 .

[29]  L. Breiman.  Random Forests , 2001, Encyclopedia of Machine Learning and Data Mining.

[30]  Lorenzo Bruzzone,et al.  An extension of the Jeffreys-Matusita distance to multiclass cases for feature selection , 1995, IEEE Trans. Geosci. Remote. Sens..

[31]  Russell G. Congalton,et al.  A review of assessing the accuracy of classifications of remotely sensed data , 1991 .

[32]  J. R. Landis,et al.  The measurement of observer agreement for categorical data. , 1977, Biometrics.

[33]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[34]  Hanchao Zhang,et al.  Extraction of marsh wetland in Heilongjiang Basin based on GEE and multi-source remote sensing data , 2022, National Remote Sensing Bulletin.

[35]  Zhaoning Gong,et al.  Wetland mapping of Yellow River Delta wetlands based on multi-feature optimization of Sentinel-2 images , 2019, National Remote Sensing Bulletin.

[36]  Mario Chica-Olmo,et al.  An assessment of the effectiveness of a random forest classifier for land-cover classification , 2012 .

[37]  R. Congalton,et al.  Accuracy assessment: a user's perspective , 1986 .