Characterizing land-use classes in remote sensing imagery by shape metrics

Shape is an important aspect of the spatial attributes of land-use segments in remotely sensed imagery, but it is still rarely used as a component in land-use classification or image-based land-use analysis. This study aimed to quantitatively characterize land-use classes using shape metrics. The study was conducted in a case area located in southern China, covering 12 scenes of SPOT-5 images. There were a total of 10 metrics selected for the analysis: convexity (CONV), solidity (SOLI), elongation (ELONG), roundness (ROUND), rectangular fitting (RECT), compactness (COMP), form factor (FORM), square pixel metric (SqP), fractal dimension (FD), and shape index (SI). The last five metrics were used to measure the complexity of shape. Six land-use classes were investigated in the case area: roads; cultivated lands; settlements; rivers; ponds; and forest and grass lands. The results showed that all the typical shape properties of the land-use segments can be well measured by shape metrics. We identified the land-use classes whose values were significantly differentiated from the other classes for each metric. Finally, we selected five shape metrics (SOLI, ELONG, ROUND, RECT, FORM) by visual comparison and statistical analysis of the metrics values, and deduced the “shape metric signatures” (SMS) of the different land-use classes. SMS were found to be accurate and predictive discriminators of land-use classes within the study area. Our results showed that SMS can clearly distinguish spectrally similar land-use classes. The results of this study will help to build a more accurate and intelligent object-oriented classification system for land-use classes.