Bathymetry retrieval from optical images with spatially distributed support vector machines

This paper presents a spatially distributed support vector machine (SVM) system for estimating shallow water bathymetry from optical satellite images. Unlike the traditional global models that make predictions from a unified global model for the entire study area, our system uses locally trained SVMs and spatially weighted votes to make predictions. By using IKONOS-2 multi-spectral image and airborne bathymetric LiDAR water depth samples, we developed a spatially distributed SVM system for bathymetry estimates. The distributed model outperformed the global SVM model in predicting bathymetry from optical satellite images, and it worked well at the scenarios with a low number of training data samples. The experiments showed the localized model reduced the bathymetry estimation error by 60% from RMSE of 1.23 m to 0.48 m. Different from the traditional global model that underestimates water depth near shore and overestimates water depth offshore, the spatially distributed SVM system did not produce regional prediction bias and its prediction residual exhibited a random pattern. Our model worked well even if the sample density was much lower: The model trained with 10% of the samples was still able to obtain similar prediction accuracy as the global SVM model with the full training set.

[1]  W J Lillycrop,et al.  Airborne lidar bathymetry : The SHOALS system , 2000 .

[2]  J. Théau,et al.  Recent applications of unmanned aerial imagery in natural resource management , 2014 .

[3]  Study on Remote Sensing of Water Depths Based on BP Artificial Neural Network , 2007 .

[4]  Anthony M. Filippi,et al.  Geographically Adaptive Inversion Model for Improving Bathymetric Retrieval From Satellite Multispectral Imagery , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[5]  Hassan Mohamed,et al.  Bathymetry Determination from High Resolution Satellite Imagery Using Ensemble Learning Algorithms in Shallow Lakes : Case Study , 2022 .

[6]  Jay Gao.  Bathymetric mapping by means of remote sensing : methods , accuracy and limitations , 2009 .

[7]  A. Hegde,et al.  Bathymetry Mapping Using Landsat 8 Satellite Imagery , 2015 .

[8]  Paul Harris,et al.  Spatial Prediction of Coastal Bathymetry Based on Multispectral Satellite Imagery and Multibeam Data , 2015, Remote. Sens..

[9]  Adrian Baddeley,et al.  Spatial Point Processes and their Applications , 2007 .

[10]  Andrew G. Birt,et al.  Computer-based synthetic data to assess the tree delineation algorithm from airborne LiDAR survey , 2011, GeoInformatica.

[11]  Xiuping Jia,et al.  Spatially Constrained Multiple Endmember Spectral Mixture Analysis for Quantifying Subpixel Urban Impervious Surfaces , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[12]  Sarah M. Hamylton,et al.  Derivation of High-Resolution Bathymetry from Multispectral Satellite Imagery: A Comparison of Empirical and Optimisation Methods through Geographical Error Analysis , 2015, Remote. Sens..

[13]  Subhash C. Bagui,et al.  Combining Pattern Classifiers: Methods and Algorithms , 2005, Technometrics.

[14]  Qunying Huang,et al.  Channel bar feature extraction for a mining-contaminated river using high-spatial multispectral remote-sensing imagery , 2016 .

[15]  Zhilei Lin,et al.  A support vector machine classifier based on a new kernel function model for hyperspectral data , 2016 .

[16]  Ludmila I. Kuncheva,et al.  Combining Pattern Classifiers: Methods and Algorithms , 2004 .

[17]  Shan Liu,et al.  Performance of two neural network models in bathymetry , 2015 .

[18]  Victor F. Rodriguez-Galiano,et al.  Evaluation of different machine learning methods for land cover mapping of a Mediterranean area using multi-seasonal Landsat images and Digital Terrain Models , 2014, Int. J. Digit. Earth.

[19]  J. Campbell.  Introduction to remote sensing , 1987 .

[20]  Jared Kibele,et al.  Nonparametric Empirical Depth Regression for Bathymetric Mapping in Coastal Waters , 2016, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[21]  S. Dark,et al.  The modifiable areal unit problem (MAUP) in physical geography , 2007 .

[22]  Javier Marcello,et al.  High-Resolution Maps of Bathymetry and Benthic Habitats in Shallow-Water Environments Using Multispectral Remote Sensing Imagery , 2015, IEEE Transactions on Geoscience and Remote Sensing.

[23]  Le Wang,et al.  An object-based SVM method incorporating optimal segmentation scale estimation using Bhattacharyya Distance for mapping salt cedar (Tamarisk spp.) with QuickBird imagery , 2015 .

[24]  D. Lyzenga.  Passive remote sensing techniques for mapping water depth and bottom features. , 1978, Applied optics.

[25]  R. Stumpf,et al.  Determination of water depth with high‐resolution satellite imagery over variable bottom types , 2003 .

[26]  Qiusheng Wu,et al.  Prediction of Water Depth From Multispectral Satellite Imagery—The Regression Kriging Alternative , 2015, IEEE Geoscience and Remote Sensing Letters.

[27]  Robert Frouin,et al.  Water Column Correction for Coral Reef Studies by Remote Sensing , 2014, Sensors.

[28]  C. Deng,et al.  A spatially adaptive spectral mixture analysis for mapping subpixel urban impervious surface distribution , 2013 .

[29]  Shu Gao,et al.  Shallow Water Depth Inversion Based on Data Mining Models , 2015 .

[30]  L. S. Galvão,et al.  Potential of multispectral and hyperspectral data to detect saline-exposed soils in Brazil , 2015 .