Texture analysis using textons

Datum konání: 08.10.2014
Přednášející: Murk Bottema
Odpovědná osoba: Zitová

Many forms of texture analysis have been used as part of various image analysis tasks such as segmentation, recognition, detection, and characterisation of images or object in images. A relatively recent addition to the repertoire of texture methods is the idea of textons. The method of textons is distinguished by the use of clustering to determine commonly occurring local texture patterns. Here, the basic notion of textons and various extensions and modifications will be presented along with applications to illustrate implementation. The applications are drawn from our current research projects and include breast cancer risk assessment and the characterisation of trabecular bone