Abstract
The recognition of textures involves multiple local filtering and subsequent averaging of the image to produce a feature space in which textures are represented as cluster centers. Results are presented for a two variate algorithm involving the features “brightness” and “roughness”. Both regular and random textures generated by program are used. The final result of classification is expressed in terms of the original textures: the computer “paints” what it has seen — and makes rather understandable errors at the borders between different textures.
Similar content being viewed by others
Literatur
Rosenfeld,A., Thurston,M.: Visual texture analysis. Technical Report TR-70-129, Computer Science Center, University of Maryland, 1970.
Rosenfeld,A., Thurston,M., Lee,Y.-H.: Edge and curve detection: Further Experiments, IEEE Transactions on Computers, July 1972.
Johnson,G.E., Lipkin,B.S., Rosenfeld,A., Lee,Y.-H.: Some studies in texture discrimination. Technical Report TR-183, Computer Science Center, University of Maryland, 1970.
Sutton,R.N., Hall,E.L.: Texture measures for automatic classification of pulmonary disease. IEEE Transactions on computers, July 1972.
Alparslan Akca: Die Verwendung quantitativ erfaßter Merkmale der photographischen Textur von Schwarz-Weiß-Luftbildern für die Identifizierung einiger Objekte. Allgemeine Forst- und Jagdzeitung (1971).
Triendl,E.: Darstellung bildlicher Information durch synthetische Texturen. Bildmessung und Luftbildwesen, 4/1972.
Triendl,E.: Automatic terrain mapping by texture recognition. Proceedings of the 7th international symposium on remote sensing of environment, Sept. 1972.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Triendl, E.E. Texturerkennung und Texturreproduktion. Kybernetik 13, 1–5 (1973). https://doi.org/10.1007/BF00289105
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF00289105