Evaluating pixel vs. segmentation based classifiers with height differentiation on SPOT 6 imagery for urban land cover mapping

Abstract

The identification, extraction, classification and mapping of detailed, but reliable land use or land cover (LULC) data play an increasingly important role in informed decision-making whether employed in urban planning and civil engineering, intensive agriculture, the natural and environmental sciences, for example. One way of extracting LULC information is through the use of algorithms that classify multispectral satellite images according to the required standard and user legend. The meaningful classification of heterogeneous urban and city landscapes however remains challenging and is usually performed using semi-automated pixel-based, object-based, or a hybrid classification workflows. With the prevailing remote sensing (RS) technologies enabling professionals to integrate data from various sources to improve the quality of LULC classification nowadays, it negated the dependency on multispectral data alone. This study sought to explore how successful can a single-acquisition pansharpened SPOT 6 image be deconstructed into obtaining primary and secondary LULC classes using a comparison of the pixel-based versus segmentation-based classifier, performed over Soshanguve Township, South Africa. The study further assessed the effect of integrating LiDAR derived 3D land surface data into both classification processes as opposed to not at all. A supervised Maximum Likelihood classifier was executed for the pixel-based routine and the ERDAS IMAGINE Objective Tool was used for the segmentation-based approach. A total of nine LULC classes were successfully identified from the classification. The results showed that the segmentation-based approach outperformed the pixel-based approach, yet when integrating height information both segmentation and pixel-based overall accuracies increased from 67.5% to 78.8 and 57.5% to 73.8%, respectively.