Assessing the Positional and Thematic Accuracy of Remotely Sensed Data

Classification of remotely sensed imagery gives variable and often poor quality results. The causes and nature of these errors have been the subject of extensive research.. There are two different components of accuracy in the context of remote sensing: positional and thematic accuracy. Positional accuracy determines how closely the position of discrete objects shown on a rectified image (map) or in a spatial database agree with the true position on the ground, while thematic accuracy refers to the non-positional characteristic of a spatial data entity, the so-called attributes (which are derived from radiometric information). Quality control of cartographic products is usually accomplished by computing the discrepancy between well defined points present in one cartographic document with the corresponding points observed in the field, using a technique that guarantees sufficient accuracy for the analysis. In spatial databases generated from remotely sensed data, it is equally necessary the knowledge of discrepancies (errors). However, in some imagery where the number of control points is not sufficient, or where their spatial distribution is suboptimal, the use of generic features (such as roads, edges, polygons, etc) is an important alternative. The aim of this paper is to review standard methods for assessing the quality control of cartographic products in the context of remote sensing. A further aim is to present a methodology to assess the positional accuracy of spatial databases using generic features (and their spatial distribution) within the image in the validation phase, and also describe a methodology that specifically takes into account spatial pattern of errors of omission and commission, in order to present the user with an indication of the reliability of the pixel label assignments. Results show that considerable amount of research and development needs to be accomplished before the spatial characterisation of positional ant thematic accuracy associated with remote sensing can be adequately reported in standardised format and legends. Several techniques for the quality control of spatial databases using generic features are adapted to the context of remote sensing, however, one drawback for some of these approaches is the difficult to obtaining homologous points in both representations. Alternative techniques to overtake such limitations are also discussed.