Abstract:Super-resolution land-cover mapping SRM is a technique for generating land-cover thematic maps with a finer spatial resolution than the input image. Linear mixture model-based SRM LSRM is applied directly to a remotely sensed image and is composed of a spatial term that integrates the land-cover spatial pattern prior information, a spectral term that assumes that the spectral signature of each mixed pixel is composed of a weighted linear sum of endmember spectral signatures within that pixel and a balance parameter that defines the weight of the spatial term. The traditional LSRM adopts an isotropic spatial autocorrelation model in the land-cover spatial term for different classes and a fixed balance parameter for the entire image, and ignores the image local properties. The class boundaries are at risk of oversmoothing and may be imprecise, and the homogeneous regions may be unsmoothed and contain speckle-like artefacts in the result. This study proposes a locally adaptive LSRM LA-LSRM that integrates image local properties to predict fine spatial resolution pixel labels. The structure tensor is applied to detect the image local information. The LA-LSRM spatial term is locally adaptive and is composed of an anisotropic spatial autocorrelation model in which the spatial autocorrelation orientations of different classes may vary. The LA-LSRM balance parameter is locally adaptive to the different regions of the image. Such parameter obtains a relatively large value when the fine-resolution pixel is located in the homogeneous region to remove speckle-like artefacts and a relatively small value when the fine-resolution pixel is at the class boundary to preserve the edge. The LA-LSRM performance was assessed using a simulated multi-spectral image, an IKONOS multi-spectral image, a hyperspectral image produced by Airborne Visible/Infrared Imaging Spectrometer and a hyperspectral image produced by reflective optics system imaging spectrometer. Results show that the homogeneous regions were smoothed, the boundaries were better preserved and the overall accuracies were increased by LA-LSRM compared with traditional LSRM in all experiments.

Multilevel T-spline Approximation for Scattered Observations with Application to Land Remote Sensing

Discontinuity-Preserving Moving Least Squares Method

Total Least Squares with Application in Geospatial Data Processing

Multiscale scattered data analysis in samplet coordinates

Assessment of the Accuracy of Terrestrial Laser Scanners in Detecting Local Surface Anomaly

Mapping ground displacement by a multiple phase difference-based InSAR approach: with stochastic model estimation and turbulent troposphere mitigation

Comparability of multi-temporal DTMs derived from different LiDAR platforms: Error sources and uncertainties in the application of geomorphic impact studies

Surface Reconstruction Method for Measurement Data with Outlier Detection by Using Improved RANSAC and Correction Parameter

An Adaptive Multilook Approach for Small Sets of Multitemporal SAR Data Based on Adaptive Joint Data Vector

Optimal Estimation of Distributed Scatterer Phase History Parameters from Meter-Resolution SAR Data.

Adaptive Hierarchical B-Spline Surface Approximation Of Large-Scale Scattered Data

Msplit Estimation with Local or Global Robustness Against Outliers—Applications and Limitations in LiDAR Data Processing

Smooth digital terrain modelling in irregular domain using finite element thin plate splines and adaptive refinement

A Mars Local Terrain Matching Method Based on 3D Point Clouds

Locally Adaptive Linear Mixture Model-Based Super-Resolution Land-Cover Mapping Based on a Structure Tensor

An Extension of Multiquadric Method Based on Trend Analysis for Surface Construction

Robust approximation of the Medial Axis Transform of LiDAR point clouds as a tool for visualisation

Sparse Approximate Solution of Fitting Surface to Scattered Points by MLASSO Model

TSSS: A Novel Triangulated Spherical Spline Smoothing for Surface-based Data

Surface reconstruction from point cloud using a semi-Lagrangian scheme with local interpolator

Accurate 3D comparison of complex topography with terrestrial laser scanner: application to the Rangitikei canyon (N-Z)