Ultrasonic flaw Signal Classification based on Curvelet Transform and Support Vector Machine
A. Pradeep Kumar
Dr. A. Pradeep Kumar, Assistant Professor, Department of Electronics and Communications Engineering, Malla Reddy Engineering College, Hyderabad, Telanagana, India.
Manuscript received on 10 December 2018 | Revised Manuscript received on 17 December 2018 | Manuscript Published on 26 December 2018 | PP: 449-453 | Volume-8 Issue- 2S2 December 2018 | Retrieval Number: ES2136017519/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This paper presents the classification of ultrasonic flaw signal with the use of curvelet transform method and support vector machine. The curvelet transform as a not merely to achieve time frequency manifestation of signal, but also to be used for curvelet signal decomposition and successive parameter assessment. Faults are detected by using a digital flaw detecting method which is considered as the primary tool to obtain the carbon fiber signals as an unbreakable polymer sample with de-lamination and de-bonding. Discrete curvelet transform can be computed ultrasonic signals in time domain by enlightening features are extracted from signals of curvelet coefficients. Finally, SVM chosen by dissimilar techniques are in use as input and train by the classifier. So the kernel function has been checking the data with combination of SVM parameters. Experimental outcome prove the validation and verification of flaw signal with curvelet transform and SVM tool, it deals with classification for ultrasonic signals utmost accurately.
Keywords: Curvelet Transform, SVM, Ultrasonic Flaw Signals, Kernel Function.
Scope of the Article: Communication