Automatic detecting and classifying defects during eddy current inspection of riveted lap-joints
ABSTRACT This article presents a novel method for automatic detection and classification of cracks located in the second lower layer of the aircraft lap-joints during Eddy Current (EC) inspection. The cracks originating from the rivet holes were detected using a tailor-made deep penetrating EC probe. The proposed method consists of three steps: pre-processing, feature extraction and classification. The pre-processing, performed before the feature extraction included median filtering, rotation and de-biasing of the EC patterns. The rotation of the patterns was performed so that energy of the responses to the rivets was maximized along the quadrature direction, while the defect responses were maximized in the in-phase direction in the impedance plane. Feature extraction was then performed using four different methods: discrete wavelet transform, Fourier descriptors, principal component analysis (PCA) and block mean values. The classification was performed using a standard multi-layer perceptron (MLP) neural network. All the pre-processing methods showed similar classification performance on the used data set, but the PCA method compressed the data best.
- [show abstract] [hide abstract]
ABSTRACT: A major concern arising from the classification of spectral data is that the number of variables or dimensionality often exceeds the number of available spectra. This leads to a substantial deterioration in performance of traditionally favoured classifiers. It becomes necessary to decrease the number of variables to a manageable size, whilst, at the same time, retaining as much discriminatory information as possible. A new and innovative technique based on adaptive wavelets, which aims to reduce the dimensionality and optimize the discriminatory information is presented. The discrete wavelet transform is utilized to produce wavelet coefficients which are used for classification. Rather than using one of the standard wavelet bases, we generate the wavelet which optimizes specified discriminant criteriaIEEE Transactions on Pattern Analysis and Machine Intelligence 11/1997; · 4.80 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: In a series of papers, Donoho and Johnstone develop a powerful theory based on wavelets for extracting nonsmooth signals from noisy data. Several nonlinear smoothing algorithms are presented which provide high performance for removing Gaussian noise from a wide range of spatially inhomogeneous signals. However, like other methods based on the linear wavelet transform, these algorithms are very sensitive to certain types of non-Gaussian noise, such as outliers. In this paper, we develop outlier resistant wavelet transforms. In these transforms, outliers and outlier patches are localized to just a few scales. By using the outlier resistant wavelet transforms, we improve upon the Donoho and Johnstone nonlinear signal extraction methods. The outlier resistant wavelet algorithms are included with the S+Wavelets object-oriented toolkit for wavelet analysis. 1 INTRODUCTION The introduction of wavelets in the late 1980's has spawned a flurry of research activity, exploring new techniques for ...02/1996;
- second 01/1990: pages 70; Academic Press.