Bitte wählen Sie ihr Lieferland und ihre Kundengruppe
Many texture-segmentation mechanisms take the form of an elaborate bank of filters. Commonly used within such mechanisms are the Gabor elementary functions (GEFs). While filter-bank-based mechanisms show promise and some analytical work has been done to demonstrate the efficacy of GEFs, the relationships between texture differences and the filter configurations required to discriminate these diferences remain largely unknown. In this paper, the authors give analytical and experimental evidence that suggests that various types of discontinuities can occur at texture boundaries when appropriately 'tuned' GEF-based filters are applied to a textured image; thus, by applying a discontinuity-detection scheme (i.e, edge detector) to such a filtered image, one can segment the image into different textured regions. Using a mathematically defined 1-D texture model, the authors show analytically that, depending on the anture of the texture difference, this output discontinuity exhibits certain characteristic signatures. If two adjacent textures differ in spatial-frequency content, this signature exhibits a step change at the location of the texture boundary. If the two adjacent textures differ only in a phase shift, the signature exhibits either a ridge or a valley at the location of the texture boundary. More complex textured images may exhibit either a combination step and ridge/valley signature at a texture boundary, or it may exhibit a step change in average local statistical variation. In the sections o follow, the authors develop this theory in detail. They begin with a discussion of GEFs and their properties and how they configure them for texture analysis. They then define a texture model and show under what conditions the various types of signatures occur. This is followed by experimental results.