In this paper, we present through the experimental study the use of support vector machines (SVMs) in the automatic classification of weld defects in radiographic images. SVM is a machine learning tool used for classification and regression and it is well known for binary classification, but there are many approaches for multiclass classification, the most popular are one versus all and one versus one. The performance of the proposed classification system is evaluated using hundreds of radiographic images representing four types of defects. The experimental results show that the SVM classifier is an efficient automatic weld defect classification algorithm and can achieve high accuracy percent and is faster than multilayer perceptron artificial neural network (MLP-ANN).

The interpretation of possible weld discontinuities in industrial radiography is ensured by human interpreters. Consequently, it is submitted to subjective considerations such as the aptitude and the experiment of the interpreter, in addition of the poor quality of radiographic images, due essentially to the exposure conditions. These considerations make the weld quality interpretation inconsistent, labor intensive and sometimes biased. It is thus desirable to develop computer-aided techniques to assist the interpreter in evaluating the quality of the welded joints. For the characterization of the weld defect region, looking for features which are invariant regarding the usual geometric transformations proves to be necessary because the same defect can be seen from several angles according to the orientation and the distance from the welded framework to the radiation source. Thus, a set of invariant geometrical attributes which characterize the defect shape is proposed. The principal component analysis technique is used in order to reduce the number of attribute variables in the aim to give better performance for defect classification. Thereafter, an artificial neural network for weld defect classification was used. The proposed classification consists in assigning the principal types of weld defects to four categories according to the morphological characteristics of the defects usually met in practice.

In this paper, a parametric and unsupervised histogram-based image segmentation method is presented. The histogram is assumed to be a mixture of asymmetric generalized Gaussian distributions. The mixture parameters are estimated by using the Expectation Maximization algorithm. Histogram fitting and region uniformity measures on synthetic and real images reveal the effectiveness of the proposed model compared to the generalized Gaussian mixture model.