Area-based validity assessment of single- and multi-class object-based image analysis

Object-based image analysis (OBIA) produces results that have both thematic and geometric properties. A classified image contains objects that not only belong to particular classes but also have spatial properties such as location and shape. Therefore, any accuracy assessment must (but often does not) take into account both thematic and geometric properties of the classified objects. Using area-based measures to compare classified objects to corresponding reference objects, accuracy information for both thematic and geometric assessment is available. Our methods provide area-based measures with application to both a single-class feature extraction and multi-class object-based land cover analysis. This work showcases the utility of the methods for assessing the accuracy of OBIA derived classifications provided the reference data is accurate and of comparable scale.