Abstract
An algorithm is considered for constructing a system of technical flaw detection of welded joints. An image recognition theory based on stochastic geometry is the foundation of the algorithm. Basing the recognition on a large number of image tags allows for more problems to be solved, including those with multiple structure classes.
Similar content being viewed by others
REFERENCES
N. G. Fedotov, Stochastic Geometry Methods in Pattern Recognition [in Russian], Radio i Svyaz', Moscow (1990).
N. G. Fedotov and A. A. Kadryov, Avtometriya, No. 1, 88 (1996).
N. G. Fedotov and L. A. Shulga, 12th Scandanavian Conference on Image Analysis, Bergen, Norway, (2001), V. I, II, p. 373.
GOST 7512-82. Nondestructive Testing. Welded Joints. Radiographic Method.
OST 102-51-85. Nondestructive Testing. Welded Pipeline Joints. Radiographic Method.
N. G. Zagoruiko, Applied Methods of Data and Knowledge Analysis [in Russian], Izd. Inst. Matematiki, Novosibirsk (1999).
M. M. Mel'nikov, Sensors of Measuring, Monitoring, and Control Systems [in Russian], Izd. Penz. Gos. Univ., Penza (2001).
Rights and permissions
About this article
Cite this article
Fedotov, N.G., Nikiforova, T.V. Technical Flaw Detection Based on a New Image Recognition Theory. Measurement Techniques 45, 1247–1253 (2002). https://doi.org/10.1023/A:1022981006436
Issue Date:
DOI: https://doi.org/10.1023/A:1022981006436