Overall Accuracy Estimation for Geographic Object-based Image Classification
Julien Radoux*, Patric Bogaert, and Pierre Defourny
Earth and Life Institute-Environmental Sciences, Universite catholique de Louvain, Louvain-la-Neuve, Belgium
ABSTRACT: Geographic object-based image analysis is a processing method where groups of spatially adjacent pixels are classified as elementary units. This approach raises concerns about the design of subsequent validation strategies. Indeed, classical point-based sampling strategies based on the spatial distribution of sample points (using systematic, probabilistic or stratified probabilistic sampling) do not rely on the same concept of objects. New methods explicitly built on the concept of objects used for the classification step are thus needed. An original object-based sampling strategy is therefore proposed and compared with other approaches used in the literature for the thematic accuracy assessment of object-based classifications. The new sampling scheme and sample analysis are founded on a sound theoretical framework based on few working hypotheses. The performance of the sampling strategies is quantified using object-based classifications results simulated for a Quickbird imagery. The bias and the variance of the overall accuracy estimates were used as indicators of the methods benefits. The main advantage of the object-based overall accuracy predictor is its performance: for a given confidence interval, it requires less sampling units than the other methods. In many cases, this can help to noticeably reduce the sampling effort. The use of objects-based sampling units leads to practical and conceptual issues, which are sometimes, but not always, similar to those of point-based accuracy assessment. These issues (mixed entities, spatial correlation, effect of the geolocation errors, sample representativity) are discussed with regard to the representation of environmental variables together with the limitations of the proposed method.
Keywords: Overall accuracy; Object; Spatial region