Sub-pixel mapping: a comparison of techniques.

Remotely sensed images often contain a combination of both pure and mixed pixels. Hard classification techniques assign mixed pixels to the class with the highest proportion of coverage or probability. Loss of information is inevitable during this process. Soft classification techniques were introduced to correct for this loss: they assign pixel fractions to the land cover classes corresponding to the represented area inside a pixel. The assignment of fractions to the different classes however, renders no information about the location of these fractions inside the pixel. Atkinson (1997) stated that it is possible to assign the fractions spatially to so called ’subpixels’. Every pixel is thus divided into a predefined number of sub-pixels, allowing a more spatially detailed representation of the lower resolution pixels. Sub-pixel mapping algorithms have been applied in a variety of forms and on fraction images of varying spatial resolutions, but all algorithms share one common property: accuracy assessment of sub-pixel mapping algorithms is impossible because of missing high resolution ground truth imagery. Therefore, high resolution reference classifications are degraded to yield artificial fraction images. When these artificial fraction images are used as input to the sub-pixel mapping process, the original reference image can serve as ground truth. This way, accuracy assessment of sub-pixel mapping algorithms is facilitated. The aim of this work is to use identical reference images for different sub-pixel mapping techniques, allowing for comparison of the performance of different techniques.

[1]  Jacob Cohen.  A Coefficient of Agreement for Nominal Scales , 1960 .

[2]  Bernard De Baets,et al.  Direct sub-pixel mapping exploiting spatial dependence , 2004, IGARSS 2004. 2004 IEEE International Geoscience and Remote Sensing Symposium.

[3]  Hugh G. Lewis,et al.  Super-resolution land cover pattern prediction using a Hopfield neural network , 2002 .

[4]  G. Foody.  Sharpening fuzzy classification output to refine the representation of sub-pixel land cover distribution , 1998 .

[5]  Lieven Verbeke,et al.  Sub-pixel mapping and sub-pixel sharpening using neural network predicted wavelet coefficients , 2004 .

[6]  Paul Aplin,et al.  Sub-pixel land cover mapping for per-field classification , 2001 .

[7]  Robert De Wulf,et al.  Land cover mapping at sub-pixel scales using linear optimization techniques , 2002 .

[8]  Christopher Jennison,et al.  A Subpixel Image Restoration Algorithm , 1997 .

[9]  Hugh G. Lewis,et al.  Super-resolution mapping of urban scenes from IKONOS imagery using a Hopfield neural network , 2001, IGARSS 2001. Scanning the Present and Resolving the Future. Proceedings. IEEE 2001 International Geoscience and Remote Sensing Symposium (Cat. No.01CH37217).

[10]  John R. Schott,et al.  Application of Spectral Mixture Analysis and Image Fusion Techniques for Image Sharpening , 1998 .

[11]  G. M. Foody.  The role of soft classification techniques in the refinement of estimates of ground control point location , 2002 .

[12]  F. J. A. López,et al.  Restoring SPOT images using PSF-derived deconvolution filters , 2002 .

[13]  L. P. C. Verbeke,et al.  Using genetic algorithms in sub-pixel mapping , 2003 .