A computer assisted method for image feature extraction and analysis based on B-spline algorithm

Autor: Marchetti A., Guo Y., Azzarelli L., Chimenti M.
Jazyk: angličtina
Rok vydání: 1992
Předmět:
Zdroj: Medicon '92, pp. 169–172, Capri, Italy, 5-10 July 1992
info:cnr-pdr/source/autori:Marchetti A.; Yuhong G.; Azzarelli L.; Chimenti M./congresso_nome:Medicon '92/congresso_luogo:Capri, Italy/congresso_data:5-10 July 1992/anno:1992/pagina_da:169/pagina_a:172/intervallo_pagine:169–172
VI Mediterranean Conference on Medical and Biological Engineering-Medicon 92, pp. 169–172, Capri, Italy, 5-10 July 1992
info:cnr-pdr/source/autori:Marchetti A.; Guo Y.; Azzarelli L.; Chimenti M./congresso_nome:VI Mediterranean Conference on Medical and Biological Engineering-Medicon 92/congresso_luogo:Capri, Italy/congresso_data:5-10 July 1992/anno:1992/pagina_da:169/pagina_a:172/intervallo_pagine:169–172
Popis: Biomedical image analysis for diagnostic purposes is executed by following a quite complicated process, that substantially implies to recognize some structures and to evaluate the dimensions and the visual appearance of some objects contained in different regions of an image. In order to execute this process by means of digital image processing it is necessary to define the logical process followed by a human observer for diagnostic understanding and evaluation. Images used in medicine are normally plenty of information but they are also affected by artifacts and a large amount of noise, so that it is not easy to realize fully automatic procedures for image analysis with an adequate degree of reliability. On the contrary it results to be useful to develop interactive procedures for image evaluation: these procedures allow to support the human task, especially in routinary operations and numerical data can be easily extracted from images in an objective way. The described method is implemented in an interactive procedure that requests a limited control by the user and allows to obtain a large set of parametric measurements of different features.
Databáze: OpenAIRE