Multispectral and Panchromatic Data Fusion Assessment Without Reference

This paper introduces a novel approach for evaluating the quality of pansharpened multispectral (MS) imagery without resorting to reference originals. Hence, evaluations are feasible at the highest spatial resolution of the panchromatic (PAN) sensor. Wang and Bovik’s image quality index (QI) provides a statistical similarity measurement between two monochrome images. The QI values between any couple of MS bands are calculated before and after fusion and used to define a measurement of spectral distortion. Analogously, QI values between each MS band and the PAN image are calculated before and after fusion to yield a measurement of spatial distortion. The rationale is that such QI values should be unchanged after fusion, i.e., when the spectral information is translated from the coarse scale of the MS data to the fine scale of the PAN image. Experimental results, carried out on very high-resolution Ikonos data and simulated Pléiades data, demonstrate that the results provided by the proposed approach are consistent and in trend with analysis performed on spatially degraded data. However, the proposed method requires no reference originals and is therefore usable in all practical cases. Introduction Remote sensing image fusion techniques aim at integrating the information conveyed by data acquired with different spatial and spectral resolution from satellite or aerial platforms (Wald, 1999). The main goal is photo analysis, but also automated tasks such as feature extraction and segmentation/classification have been found to benefit from fusion (Colditz et al., 2006; Bruzzone et al., 2006). A variety of image fusion techniques are devoted to merge multispectral (MS) and panchromatic (PAN) images, which exhibit complementary characteristics of spatial and spectral resolutions (Wang et al., 2005). Such an application of data fusion is often called pansharpening. Injection in the resampled MS images of spatial details extracted from the PAN image has been found to be adequate for preserving the spectral characteristics (Chavez Jr. et al., 1991). MultiMultispectral and Panchromatic Data Fusion Assessment Without Reference Luciano Alparone, Bruno Aiazzi, Stefano Baronti, Andrea Garzelli, Filippo Nencini, and Massimo Selva resolution analysis, based on undecimated wavelets decompositions and Laplacian pyramids, has proven itself effective to implement fusion at different resolutions (Núñez et al., 1999; Aiazzi et al., 2002a). Quantitative results of data fusion are provided thanks to the availability of reference originals obtained either by simulating the target sensor by means of high-resolution data from an airborne platform (Laporterie-Déjean et al., 2005), or by degrading all available data to a coarser resolution and carrying out fusion from such data. In practical cases, this strategy is not feasible. The underlying assumption, however, is that fusion performances are invariant to scale changes (Wald et al., 1997). Hence, algorithms optimized to yield best results at coarser scales, i.e., on spatially degraded data, should still be optimal when the data are considered at finer scales, as it happens in practice. This assumption may be reasonable in general, but unfortunately may not hold for very high-resolution data, especially in a highly detailed urban environment, unless the spatial degradation is performed by using lowpass filters whose frequency responses match the shape of the modulation transfer functions (MTF) of the sensor (Aiazzi et al., 2006). In this work, we present a global index capable of measuring the quality of pansharpened MS images and working at the full scale without performing any preliminary degradation of the data. The spatial and spectral distortions are separately calculated from the fused MS image, the source MS image, and the PAN image. A combination of the spectral and spatial distortion indices may be performed to obtain a unique quality index. The rationale is that the interrelationships between any couple of spectral bands and between each band and the PAN image should be unchanged after fusion. Changes in the former are responsible for spectral distortion. Changes in the latter indicate spatial distortion. The underlying assumption of inter-scale preservation of cross-similarity is demonstrated by the fact that the true high-resolution MS data, whenever available, exhibit spectral and spatial distortions that are both zero, within the approximations of the model, and definitely lower than those attained by any fusion method. A thorough experimental section highlights the assets of the proposed approach compared with the two main approaches in the literature. The paper is organized as follows. The two major protocols for MS PAN image fusion assessment and the related statistical indices are reviewed next, followed by a description of how the mutual relationships among the PHOTOGRAMMETRIC ENGINEER ING & REMOTE SENS ING Feb r ua r y 2008 193 Luciano Alparone is with the Department of Electronics & Telecommunications, University of Florence, 3 Via Santa Marta, 50139 Florence, Italy ([email protected]). Bruno Aiazzi, Stefano Baronti, and Massimo Selva are with the Institute of Applied Physics “Nello Carrara” IFAC-CNR, 10 Via Madonna Del Piano, 50019 Sesto F.no (Florence), Italy. Andrea Garzelli and Filippo Nencini are with the Department of Information Engineering, University of Siena, 56 Via Roma, 53100 Siena, Italy. Photogrammetric Engineering & Remote Sensing Vol. 74, No. 2, February 2008, pp. 193–200. 0099-1112/08/7402–0000/$3.00/0 © 2008 American Society for Photogrammetry and Remote Sensing M-21.qxd 1/10/08 1:42 AM Page 193