Segmentation Evaluation for Object Based Remotely Sensed Image Analysis

Object based image analysis (OBIA) is a relatively new form of remote sensing which aims to overcome the failings of traditional pixel based techniques at providing accurate land-use classification for high resolution data. The failure of pixel based techniques is due to the fact that these techniques are based on the assumption that individual classes contain uniform visual properties. As we increase the spatial resolution of data the intra-class variation increases and this property of class uniformity is broken leading to very poor performance (Blaschke 2003). The human visual system (HVS) can interpret high resolution data very easily and accurately. If a truly accurate and robust automated land-use classification system is to be achieved, it must draw from research in the area of cognitive psychology and attempt to model how we as humans interpret aerial imagery. This is the aim of research in the area of OBIA