Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Standardization for improved Spatio-Temporal Image Fusion

Harkaitz Goyena, Peter M. Atkinson, Unai Pérez-Goya, M. Dolores Ugarte

TL;DR

This work tackles the alignment gap between multi-sensor imagery used for Spatio-Temporal Image Fusion (STIF) by introducing two standardization strategies: a PSF-informed automatic upscaling pipeline and Anomaly Based Satellite Image Standardization (ABSIS). The PSF-based method degrades fine-resolution data to better resemble coarse-resolution imagery through PSF simulation, co-registration, and joint optimization, while ABSIS preserves fine-scale spatial patterns and transfers temporal changes via anomaly modeling and local regressions. Empirical results across Croplands and New Cairo show upscaling yields strong linear correlations between upscaled and coarse images, whereas ABSIS significantly boosts fusion accuracy, achieving up to 49.46% RMSE reduction and 78.40% improvements in spatial Edge metrics. The findings demonstrate complementary strengths: upscaling enhances cross-sensor similarity; ABSIS improves spectral and spatial fidelity in fused outputs, offering a robust, generalizable standardization framework for multi-sensor STIF tasks.

Abstract

Spatio-Temporal Image Fusion (STIF) methods usually require sets of images with matching spatial and spectral resolutions captured by different sensors. To facilitate the application of STIF methods, we propose and compare two different standardization approaches. The first method is based on traditional upscaling of the fine-resolution images. The second method is a sharpening approach called Anomaly Based Satellite Image Standardization (ABSIS) that blends the overall features found in the fine-resolution image series with the distinctive attributes of a specific coarse-resolution image to produce images that more closely resemble the outcome of aggregating the fine-resolution images. Both methods produce a significant increase in accuracy of the Unpaired Spatio Temporal Fusion of Image Patches (USTFIP) STIF method, with the sharpening approach increasing the spectral and spatial accuracies of the fused images by up to 49.46\% and 78.40\%, respectively.

Standardization for improved Spatio-Temporal Image Fusion

TL;DR

This work tackles the alignment gap between multi-sensor imagery used for Spatio-Temporal Image Fusion (STIF) by introducing two standardization strategies: a PSF-informed automatic upscaling pipeline and Anomaly Based Satellite Image Standardization (ABSIS). The PSF-based method degrades fine-resolution data to better resemble coarse-resolution imagery through PSF simulation, co-registration, and joint optimization, while ABSIS preserves fine-scale spatial patterns and transfers temporal changes via anomaly modeling and local regressions. Empirical results across Croplands and New Cairo show upscaling yields strong linear correlations between upscaled and coarse images, whereas ABSIS significantly boosts fusion accuracy, achieving up to 49.46% RMSE reduction and 78.40% improvements in spatial Edge metrics. The findings demonstrate complementary strengths: upscaling enhances cross-sensor similarity; ABSIS improves spectral and spatial fidelity in fused outputs, offering a robust, generalizable standardization framework for multi-sensor STIF tasks.

Abstract

Spatio-Temporal Image Fusion (STIF) methods usually require sets of images with matching spatial and spectral resolutions captured by different sensors. To facilitate the application of STIF methods, we propose and compare two different standardization approaches. The first method is based on traditional upscaling of the fine-resolution images. The second method is a sharpening approach called Anomaly Based Satellite Image Standardization (ABSIS) that blends the overall features found in the fine-resolution image series with the distinctive attributes of a specific coarse-resolution image to produce images that more closely resemble the outcome of aggregating the fine-resolution images. Both methods produce a significant increase in accuracy of the Unpaired Spatio Temporal Fusion of Image Patches (USTFIP) STIF method, with the sharpening approach increasing the spectral and spatial accuracies of the fused images by up to 49.46\% and 78.40\%, respectively.
Paper Structure (20 sections, 13 equations, 9 figures, 8 tables, 1 algorithm)

This paper contains 20 sections, 13 equations, 9 figures, 8 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Standardization approaches
  3. Automatic Upscaling
  4. Simulate Sentinel-3 Point Spread Function
  5. Co-registration
  6. Joint Gaussian Filter and Co-registration
  7. Sub-pixel level co-registration
  8. Generalization
  9. Anomaly-Based Satellite Image Standardization (ABSIS)
  10. Spatial Pattern Capture (SPC)
  11. Anomaly Capture and Correction (ACC)
  12. Rebuilding the corrected image
  13. Experimental Analysis
  14. Data
  15. Standardization experiments
  16. ...and 5 more sections

Figures (9)

  • Figure 1: RGB representations, from left to right, of the smoothed Sentinel-3 SYN image, aggregated Sentinel-2 and original Sentinel-2 MSI image for the croplands region.
  • Figure 2: RGB representations, from left to right, of the smoothed Sentinel-3 SYN image, aggregated Sentinel-2 and original Sentinel-2 MSI image for the New Cairo region.
  • Figure 3: Flowchart of the ABSIS method
  • Figure 4: RGB representation of the fine-scale (left) and aggregated (right) imagery for the Croplands (top) and New Cairo (bottom) datasets. Images correspond to pixel-by-pixel averages over the time-series.
  • Figure 5: RGB representations of the standardization results for the Croplands region. From left to right, the smoothed Sentinel-3 image, the pairwise upscaled image, the generalized upscaling image and the ABSIS standardized image.
  • ...and 4 more figures