Beware of per-pixel characterization of land cover

By John R. Townshend, Chengquan Huang, Satya Kalluri, Ruth S. DeFries, Shunlin Liang, K. Yang. 2000

236 citations recorded by the source corpus.

Subfields: Ecology, Global and Planetary Change, Media Technology

View Paper: https://www.semanticscholar.org/paper/213bb0297d71496813001d76f746578a7da0cc9e

How researchers see this paper

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

In Ecology

In Global and Planetary Change

In Media Technology

Citations in the corpus

Automated classification of landform elements using object-based image analysis (2006, background): “Poor transferability is also generally stated for most pixelbased remote sensing classifications (Townshend et al., 2000).”
Forest Type Mapping using Object-specific Texture Measures from Multispectral Ikonos Imagery: Segmentation Quality and Image Classification Issues (2009, background): “Pixel-based approaches also ignore the context and the spectral values of adjacent pixels (Fisher, 1997; Townshend et al., 2000; Brandtberg and Warner, 2006).”
Impact of sensor's point spread function on land cover characterization: Assessment and deconvolution (2002, methodology): “Previously, we showed in a simulation study (Townshend et al., 2000) that the impact of PSF effects in estimating subpixel land cover proportions could also be reduced simply by aggregating derived proportion images to coarser resolutions.”
A multi-scale segmentation/object relationship modelling methodology for landscape analysis (2003, background): “By ignoring concepts of hierarchy and scale in the landscape processes driving pattern creation, these approaches are still overly ‘pixel-centred’ (Townshend et al., 2000).”
Remotely sensed spectral heterogeneity as a proxy of species diversity: Recent advances and open challenges (2010, background): “pixels occupied solely byone cover type, do not accurately represent actual ecological patterns (Townshend et al., 2004).”
Development of a MODIS tree cover validation data set for Western Province, Zambia (2002, background): “Townshend et al. (2000) tested these effects and warn data users to beware making per pixel characterizations.”
What’s wrong with pixels? Some recent developments interfacing remote sensing and GIS (2001, background): “As Townshend et al. (2000) point out, a significant, but usually ignored problem with per-pixel characterisation of land cover is that a substantial proportion of the signal apparently coming from the land area represented by a pixel comes from the surrounding pixels.”