You are here:
ASSESSMENT OF LANDSCAPE CHARACTERISTICS ON THEMATIC IMAGE CLASSIFICATION ACCURACY
Citation:
Smith, J H., J D. Wickham, S. V. Stehman, AND L. Yang. ASSESSMENT OF LANDSCAPE CHARACTERISTICS ON THEMATIC IMAGE CLASSIFICATION ACCURACY. Presented at American Society of Photogrammetry and Remote Sensing 2002 Annual Meeting, Washington, DC, April 16-21, 2001.
Impact/Purpose:
Our research objectives are to: (a) develop new methods using satellite remote sensor data for the rapid characterization of LC condition and change at regional to national scales; (b) evaluate the utility of the new NASA-EOS MODIS (Moderate Resolution Imaging Spectrometer) leaf area index (LAI) measurements for regional scale application with landscape process models (e.g., biogenic emissions and atmospheric deposition); (c) provide remote sensor derived measurement data to advance the development of the next generation of distributed landscape process-based models to provide a predictive modeling capability for important ecosystem processes (e.g., nutrients, sedimentation, pathogens, etc.); and (d) integrate in situ monitoring measurement networks with UAV and satellite based remote sensor data to provide a continuous environmental monitoring capability.
Description:
Landscape characteristics such as small patch size and land cover heterogeneity have been hypothesized to increase the likelihood of misclassifying pixels during thematic image classification. However, there has been a lack of empirical evidence, to support these hypotheses. This study utilizes data gathered as part of the accuracy assessment Of the 1992 National Land Cover Data (NLCD) set to identify and quantify the impacts of landscape characteristics. Guided step-wise logistic regression procedures were utilized to assess the impacts of patch size, land cover heterogeneity and their interaction, by NLCD class for four regional data sets- The -2 Lc)g Likelihood (12LL) values were used to formally test if the presence of a specific variable resulted in a statistically significant improvement in accuracy. In addition, evaluations of individual variables were conducted by calculating the change in the odds of a correct classification given a one-unit change in the explanatory variable (odds ratios) and the value at which there was a 50% change of correctly classifying the sample (median effective level). The analyses reveal that the impact of land cover hetergeneity and patch size play a major role in determining whether a sample was correctly classified, but that these impacts varied by region for many of the land cover classes. While the relative impact of the individual impacts did vary, their overall tendencies were consistent across all of the classes and regions examined, with accuracy decreasing as patch size decreased and land cover heterogeneity increased.