Outex is a framework for empirical evaluation of texture classification and segmentation algorithms.
The framework is being constructed according to the following design principles:
Large versatile image database. The image database contains a large collection of textures, both in form of surface textures and natural scenes. The collection of surface textures exhibits well defined variations to a given reference in terms of illumination, rotation and spatial resolution. Reliable manual segmentation of natural scenes is included in the ground truth data.
A wide range of texture classification and segmentation problems. A large collection of texture classification and segmentation problems, both supervised and unsupervised, is constructed using the image database. The diversity of the surface textures provides a rich foundation for texture classification problems: in addition to ´standard´ texture classification, problems of illumination/rotation/resolution invariant texture classification, or their combinations, are also available. Different misclassification cost functions and a prior probabilities of the classes are also incorporated.
Precise problem definition with test suites. Problems are encapsulated into welldefined test suites having precise specifications of input and output data. Specifications are provided in form of general purpose text and image files, hence the user of the framework is not constrained to any given programming environment. Test suites are delivered as individual zip files, which expand to a ´standardized´ directory and file structure.
Trust. We trust other research groups in that they will upload to the site unbiased honest results, which are obtained in accordance with the given test suite specifications.
Collaborative development. We are inviting other research groups to join the effort, for the purpose of a more thorough and efficient long term development of the framework, and to aid in gaining acceptance of the framework in the research community.
Continuing maintenance and refinement. The host research group has a history of about two decades of texture analysis research and current plan is to stay in the business for at least another decade, hence the commitment to maintain and refine the framework is strong.
For more information see: Outex - New framework for empirical evaluation of texture analysis algorithms
Extended Outex texture classification test suites. Losson et al. extended Outex. They regenerated some of the gray-level test suites from Outex original color images. They extended most Outex test suites to 68 textures and also simulated noisy and blurred images in new degraded test suites. For more details about this extended Outex textures, see http://lagis-vi.univ-lille1.fr/datasets/outex.html