| |
| | Structural Pattern Recognition of Land Use |
 | | In general, the field of graphical structural pattern recognition is concerned with the development of computational approaches which can derive, represent and analyse both the intrinsic structure of a series of entities or pattern primitives, as well the extrinsic relationships which exit between them. |
 | | The pattern features derived for each region pattern primitive may take a wide variety of forms and measurement scales, although in general, they tend to describe the morphology of regions (e.g., their area, perimeter, compactness etc) as well as the spatial relationships which exist between them (e.g., adjacency, containment, distance and direction etc). |
 | | Furthermore, by combining this structural pattern recognition system with series of approaches more commonly associated with statistical pattern recognition, it has been shown that is possible to obtain a quantitative measurement on the degree of structural separability which exists between a series urban land use pattern classes using the structural information derived. |
| www.geocomputation.org /1998/43/gc_43.htm (4199 words) |
|