Test system for defect detection in construction materials with ultrasonic waves by support vector machine and neural network

This paper introduces a newly developed test system for structure inspection and evaluation of cement-based products by applying ultrasonic test with support vector machine (SVM) classifier. In other words, this paper represents a novel method based on SVM for defect detection, classification of number of defects, and identification of defect materials. With the system, pattern of ultrasonic waves for each case of specimen can be obtained from direct and indirect measurements. Machine learning algorithm called support vector machine and artificial neural network (ANN) are employed for classification and verification of the wave patterns obtained from different samples. By applying the system, the presence or absence of a defect in mortar can be identified. Moreover, the system can also classify the number of defects and identify the defect materials being inside the mortar. For classification, input features are extracted in different ways and the numbers of training sets are varied. Base on the results from SVM, the signals extracted in frequency domain gives better performance than time domain. Using a larger training set can give more satisfactory results. In this article, the methodology is explained and the classification results are discussed. The effectiveness of the developed test system is evaluated. Comparison of the classification results that obtained by between SVM and ANN classifiers is also demonstrated. This study shows that this technique based on pattern recognition has a high potential for practical inspection of concrete structure.