Application of neural networks to microstructure classification with the use of image features

• Autorzy: Pamuła T.
• Języki publikacji: EN

Abstrakty

EN

The paper presents further results of research carried out on the application of neural networks to microstructure analysis. As presented in [1] the aim is to classify microstructures. The current approach differs in that, the inputs of neural networks are not image pixel values but image features calculated over patches of the whole image. In the course of research a set of 7 features was named, which enabled effective analysis. Appropriate training sets were prepared and different net structures tested. The results may be applied in examining the quality of metal implants used in medicine.

Słowa kluczowe

EN

image processing; microstructures; image features

PL

przetwarzanie obrazu; mikrostruktury; cechy obrazu

• Wydawca: University of Silesia, Institute of Informatics, Computer Systems Department
• Czasopismo: Journal of Medical Informatics & Technologies
• Rocznik: 2001
• Tom: Vol. 2, Part. 2
• Strony: SN39--SN43
• Opis fizyczny: Bibliogr. 5 poz., rys., tab.

Twórcy

autor

Pamuła T.

• Institute of Transport, Silesian Technical University, Katowice, ul. Krasińskiego 8

Bibliografia

• [1] PAMULA T., Neuron classifier of image objects, Journal of Medical Informatics and Technologies, Vol. 5, November 2000, pp IT-77-82, Institute of Informatics, Silesian Technical University.
• [2] BHADESHIA H.K.D.H., Neural Networks in Materials Science, ISIJ International 1999.
• [3] OSOWSKI S., Sieci neuronowe w ujęciu algorytmicznym, Warszawa, WNT, 1996.
• [4] MATLAB, v. 5.3 with Neural Network TOOLBOX, Mathworks 1999.
• [5] DUCH W., KORBICZ J.,RUTKOWSKI L., TADEUSIEWICZ R., Biocybernetyka i Inżynieria Biomedyczna, Sieci Neuronowe, Tom 6, PAN, AOW EXIT, Warszawa 2000.