Comparison of Pansharpening Algorithms: Outcome of the 2006 GRS-S Data-Fusion Contest

By Luciano Alparone, Lucien Wald, Jocelyn Chanussot, Claire Thomas, Paolo Gamba, Lori Mann Bruce. IEEE Transactions on Geoscience and Remote Sensing, 2007

912 citations recorded by the source corpus.

In January 2006, the Data Fusion Committee of the IEEE Geoscience and Remote Sensing Society launched a public contest for pansharpening algorithms, which aimed to identify the ones that perform best. Seven research groups worldwide participated in the contest, testing eight algorithms following different philosophies [component substitution, multiresolution analysis (MRA), detail injection, etc.]. Several complete data sets from two different sensors, namely, QuickBird and simulated Pleiades, were delivered to all participants. The fusion results were collected and evaluated, both visually and objectively. Quantitative results of pansharpening were possible owing to the availability of reference originals obtained either by simulating the data collected from the satellite sensor by means of higher resolution data from an airborne platform, in the case of the Pleiades data, or by first degrading all the available data to a coarser resolution and saving the original as the reference, in the case of the QuickBird data. The evaluation results were presented during the special session on data fusion at the 2006 international geoscience and remote sensing symposium in Denver, and these are discussed in further detail in this paper. Two algorithms outperform all the others, the visual analysis being confirmed by the quantitative evaluation. These two methods share the same philosophy: they basically rely on MRA and employ adaptive models for the injection of high-pass details.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/6fa0cec7af837349f234f38e334aaa86c517f731

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

Detail Injection-Based Deep Convolutional Neural Networks for Pansharpening (2021, methodology): “Geoscience and Remote Sensing Society in 2006 [3], [4] and”
Multi-source remote sensing data fusion: status and trends (2010, methodology): “The best algorithm was identified through both visual and quantitative assessment (Alparone et al. 2007).”
Multimodal Classification of Remote Sensing Images: A Review and Future Directions (2015, background): “A wide variety of image fusion methods have been proposed in the remote sensing literature [7], [17], [18], [38], [39], being pan-sharpening methods [24], [32], [35], [40] among the most widespread.”
Hyperspectral and LiDAR Data Fusion: Outcome of the 2013 GRSS Data Fusion Contest (2014, background): “The committee organizes an annual Data Fusion Contest open to the broader community with the goal of encouraging new techniques (and evaluating existing ones) for remote sensing problems using data from different sources [21]–[25].”
Synthesis of Multispectral Images to High Spatial Resolution: A Critical Review of Fusion Methods Based on Remote Sensing Physics (2008, methodology): “THOMAS et al.: SYNTHESIS OF MULTISPECTRAL IMAGES TO HIGH SPATIAL RESOLUTION 1311 other methods [ 53 ].”
Deep Hyperspectral Image Sharpening (2018, methodology): “To acquire an HR-HSI, one class of methods combine an LR MSI with an HR panchromatic (PAN) image (gray image), which is called as the pan sharpening [30].”
Hyperspectral Pansharpening: A Review (2015, background): “The huge interest of the community on this topic is evidenced by the existence of sessions dedicated to this topic in the most important remote sensing and Earth observation conferences as well as by the launch of public contests, of which the one sponsored by the data fusion committee of the IEEE G”
Remote Sensing Image Fusion via Sparse Representations Over Learned Dictionaries (2013, methodology): “Recently, other popular multiresolution-analysisbased methods have been proposed, such as the ATWT with context-based decision (CBD) injection model [10], support value transform (SVT) [11], and Laplacian pyramids [12].”
Hyperspectral Remote Sensing Data Analysis and Future Challenges (2013, background): “fusion of multispectral images with a high spatial resolution panchromatic image [31] are eas-”
Deep learning in remote sensing: a review (2017, background): “Traditionally, the field was dominated by works fusing multispectral data with panchromatic bands [117], but more recently it has been extended to thermal [118] or hyperspectral images [119].”
A New Adaptive Component-Substitution-Based Satellite Image Fusion by Using Partial Replacement (2011, methodology): “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].”
A Critical Comparison Among Pansharpening Algorithms (2015, methodology): “It is commonly known as MTF-GLP with contextbased decision (MTF-GLP-CBD) [8], since the injection coefficient can be locally optimized by patching the image in nonoverlapping zones.”
Fusing Hyperspectral and Multispectral Images via Coupled Sparse Tensor Factorization (2018, background): “Pan-sharpening [9] is a class of seminal fusion methods which infer a high spatial resolution multispectral image from a low spatial resolution version and a panchromatic (PAN) image (gray image).”
A Convex Formulation for Hyperspectral Image Superresolution via Subspace-Based Regularization (2014, background): “A related problem that has been extensively studied is pansharpening, which addresses the fusion of MSIs and panchromatic images (PANs), the latter of which are single-band images usually covering the visible and the near-infrared spectral ranges [4]–[6].”
Optimal MMSE Pan Sharpening of Very High Resolution Multispectral Images (2008, result): “We have referred to the results of the 2006 IEEE Data Fusion Contest [25]–[27], thus comparing to the two algorithms CBD [9] and proportional additive wavelet to the L component (AWLP) [28] which have been awarded as the two most distinguished algorithms [25].”