Optimal MMSE Pan Sharpening of Very High Resolution Multispectral Images

By Andrea Garzelli, Filippo Nencini, Luca Capobianco. IEEE Transactions on Geoscience and Remote Sensing, 2008

558 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/b55e421117b56b877a4e44c9abbd3a2802122e87

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

A Multiscale and Multidepth Convolutional Neural Network for Remote Sensing Imagery Pan-Sharpening (2017, methodology): “minimizing a loss function with the prior constraints, such as the minimum mean square error based band-dependent spatial detail model [12], nonlocal optimization based on k-means clustering”
Detail Injection-Based Deep Convolutional Neural Networks for Pansharpening (2021, methodology): “the MS image interpolation using a polynomial kernel with 23 coefficients (EXP ) [17], the GS sharpening approach [12], the SFIM [13], the PRACS approach [11], the BDSD method [9], the robust BDSD approach (BDSD-PC) [10], the GLP with MTF-matched filter [58] and multiplicative injection model [59] [”
A Critical Comparison Among Pansharpening Algorithms (2015, methodology): “Due to the lack of a unique transform for extracting the component most suited for substitution, methods based on its adaptive estimation have been proposed; they are known as adaptive CS [43], [49].”
A Framework for Automatic and Unsupervised Detection of Multiple Changes in Multitemporal Images (2012, methodology): “To this purpose, we applied the minimum-mean-squareerror pansharpening method [45] to the panchromatic channel and the four bands of the multispectral images.”
PanNet: A Deep Network Architecture for Pan-Sharpening (2017, methodology): “We compare with six widely used pan-sharpening methods: PRACS [8], Indusion [19], PHLP [18], BDSD [13], SIRF [6, 7] and PNN [21].”
Contrast and Error-Based Fusion Schemes for Multispectral Image Pansharpening (2014, background): “In general the pansharpening is carried out by injecting the spatial details contained in P (and not resolved by the MS) into an upsampled version of MS [1], [2].”
A New Benchmark Based on Recent Advances in Multispectral Pansharpening: Revisiting Pansharpening With Classical and Emerging Pansharpening Methods (2021, methodology): “Close-ups of the main fusion results for the RR GE-1 Collazzone (selected bands: red, green, and blue): (a) GT, (b) EXP, (c) BT-H, (d) BDSD, (e) C-BDSD, (f) BDSD-PC, (g) C-GSA, (h) MTF-GLP-FS, (i) MTF-GLP-HPM-H, (j) MTF-GLP-HPM-R, (k) C-MTF-GLP-CBD, (l) MF, (m) FE-HPM, (n) SR-D, (o) PWMBF, (p) TV, (”
Remote Sensing Image Fusion Based on Two-Stream Fusion Network (2017, methodology): “…trous wavelet transform with the injection model 3 (ATWT M3) [9], generalized Laplacian pyramid with MTF-matched ﬁlter and multiplicative injection model (MTF GLP HPM) [64], band-dependent spatial-detail with lo-cal parameter estimation (BDSD) method [66], and CNN based pan-sharpening (PNN) [23].”
Target-Adaptive CNN-Based Pansharpening (2017, methodology): “Besides our CNN-based methods, we consider a number of well-known conventional techniques selected because of their good performance, and specifically, PRACS [8], Indusion [13], AWLP [12], ATWT-M3 [10], MTF-GLP-HPM [15], BDSD [9], and its recent extension C-BDSD [22].”
A New Pan-Sharpening Method Using a Compressed Sensing Technique (2011, methodology): “Also, as a popular idea, the inverse-problem-based methods are used to restore the original high-resolution MS image from its degraded versions, i.e., the PAN and MS images [21], [ 22 ].”
Pansharpening by Convolutional Neural Networks (2016, methodology): “In [25], a different MMSE-optimal detail image is extracted for each MS band, by evaluating band-dependent generalized intensities.”
A New Adaptive Component-Substitution-Based Satellite Image Fusion by Using Partial Replacement (2011, background): “Based on this hypothesis, various statistical evaluation indicators of spatially degraded data, such as correlation coefficient, erreur relative globale adimensionnelle de synthese, spectral angle mapper, and Q4 index, are proposed [20], [37]– [40].”
Super-Resolution of Sentinel-2 Images: Learning a Globally Applicable Deep Neural Network (2018, unknown): “As discussed earlier, there is a conceptual difference between multi-spectral super-resolution and classical pan-sharpening, in that the latter simply “copies” high-frequency information from an overlapping or nearby high-14 (Choi et al., 2011), MTF-GLP-HPM-PP (Lee and Lee, 2010) and BDSD (Garzelli ”