Multivariate time series (MTS) datasets broadly exist in numerous fields, including health care, multimedia, finance, and biometrics. How to classify MTS accurately has become a hot research topic ...since it is an important element in many computer vision and pattern recognition applications. In this paper, we propose a Mahalanobis distance-based dynamic time warping (DTW) measure for MTS classification. The Mahalanobis distance builds an accurate relationship between each variable and its corresponding category. It is utilized to calculate the local distance between vectors in MTS. Then we use DTW to align those MTS which are out of synchronization or with different lengths. After that, how to learn an accurate Mahalanobis distance function becomes another key problem. This paper establishes a LogDet divergence-based metric learning with triplet constraint model which can learn Mahalanobis matrix with high precision and robustness. Furthermore, the proposed method is applied on nine MTS datasets selected from the University of California, Irvine machine learning repository and Robert T. Olszewski's homepage, and the results demonstrate the improved performance of the proposed approach.
Automated optical inspection (AOI) has been widely used in industrial Quality Assurance (QA) procedures. Multi-task inspection in high-speed AOI systems is becoming a significant problem in the ...design. In this paper, the design of an AOI system for E-shaped magnetic core elements is briefly described and several novel algorithms are proposed to realize defects detection by this system. First, this paper proposes a robust k-tSL-center clustering method to classify the interfaces of the element into normal and damaged areas. Second, a modified Active Shape Model (ASM) method is adopted to perform shape distortion detection in real-time. Performance evaluations are carried out on an E-shaped Magnetic Core Image Database, in which all images are captured by the designed AOI system. Experimental results show that the proposed methods are more efficient, robust and accurate than state-of-the-art methods in this application.
In this paper, we describe a novel algorithm for unsupervised segmentation of images with low depth of field (DOF). First of all, a multi-scale reblurring model is used to detect the object of ...interest (OOI) in saliency space. Then, to determine the boundary of OOI, an active contour model based on hybrid energy function is proposed. In this model, a global energy item related with the saliency map is adopted to find the global minimum, and a local energy term regarding the low DOF image is used to improve the segmentation precision. In addition, an adaptive parameter is attached to this model to balance the weight of global and local energy. Furthermore, an unsupervised curve initialization method is designed to reduce the number of evolution iterations. Finally, we conduct experiments on various low DOF images, and the results demonstrate the high robustness and precision of the proposed approach.
Iris authentication is one of the most successful applications in video analysis and image processing. In this paper, several novel approaches are proposed to improve the overall performance of iris ...recognition systems. First, this paper proposes a new eyelash detection algorithm based on directional filters, which achieves a low rate of eyelash misclassification. Second, a multiscale and multidirection data fusion method is introduced to reduce the edge effect of wavelet transformation produced by complex segmentation algorithms. Finally, an iris indexing method on the basis of corner detection is presented to accelerate exhausted the 1: N search in a huge iris database. The performance evaluations are carried out on two popular iris databases, and the test results are experimentally more robust and accurate with less elapsed time compared with most existing methods.
Fault detection is fundamental to many industrial applications. With the development of system complexity, the number of sensors is increasing, which makes traditional fault detection methods lose ...efficiency. Metric learning is an efficient way to build the relationship between feature vectors with the categories of instances. In this paper, we firstly propose a metric learning-based fault detection framework in fault detection. Meanwhile, a novel feature extraction method based on wavelet transform is used to obtain the feature vector from detection signals. Experiments on Tennessee Eastman (TE) chemical process datasets demonstrate that the proposed method has a better performance when comparing with existing methods, for example, principal component analysis (PCA) and fisher discriminate analysis (FDA).
Process monitoring and fault diagnosis (PM-FD) has been an active research field since it plays important roles in many industrial applications. In this paper, we present a novel data-driven fault ...diagnosis algorithm which is based on the multivariate dynamic time warping measure. First of all, we propose a Mahalanobis distance based dynamic time warping measure which can compute the similarity of multivariate time series (MTS) efficiently and accurately. Then, a PM-FD framework which consists of data preprocessing, metric learning, MTS pieces building, and MTS classification is presented. After that, we conduct experiments on industrial benchmark of Tennessee Eastman (TE) process. The experimental results demonstrate the improved performance of the proposed algorithm when compared with other classical PM-FD classical methods.
How to select and weigh features has always been a difficult problem in many image processing and pattern recognition applications. A data-dependent distance measure can address this problem to a ...certain extent, and therefore an accurate and efficient metric learning becomes necessary. In this paper, we propose a LogDet divergence-based metric learning with triplet constraints (LDMLT) approach, which can learn Mahalanobis distance metric accurately and efficiently. First of all, we demonstrate the good properties of triplet constraints and apply it in LogDet divergence-based metric learning model. Then, to deal with high-dimensional data, we apply a compressed representation method to learn, store, and evaluate Mahalanobis matrix efficiently. Besides, a dynamic triplets building strategy is proposed to build a feedback from the obtained Mahalanobis matrix to the triplet constraints, which can further improve the LDMLT algorithm. Furthermore, the proposed method is applied to various applications, including pattern recognition, facial expression recognition, and image retrieval. The results demonstrate the improved performance of the proposed approach.
Large data sets classification is widely used in many industrial applications. It is a challenging task to classify large data sets efficiently, accurately, and robustly, as large data sets always ...contain numerous instances with high dimensional feature space. In order to deal with this problem, in this paper we present an online Logdet divergence based metric learning (LDML) model by making use of the powerfulness of metric learning. We firstly generate a Mahalanobis matrix via learning the training data with LDML model. Meanwhile, we propose a compressed representation for high dimensional Mahalanobis matrix to reduce the computation complexity in each iteration. The final Mahalanobis matrix obtained this way measures the distances between instances accurately and serves as the basis of classifiers, for example, the k-nearest neighbors classifier. Experiments on benchmark data sets demonstrate that the proposed algorithm compares favorably with the state-of-the-art methods.
In the industrial quality assurance procedures, the Automatic Visual Inspection (AVI) has been widely used for various tasks, such as dimension measurement, shape distortion detection and surface ...damage detection. First, an AVI system for E-shaped magnetic core elements is described and a surface damage inspection algorithm is proposed in this paper. Second, the paper proposed a robust K-tSL-center clustering method to improve the accuracy, robustness and efficiency of classification. Third, the gray-scale feature (S-feature) and Gabor wavelet feature (W-feature) of the interfaces of elements are extracted to combine the SW-feature and the proposed clustering method is used to classify these interfaces into normal and damaged areas. Performance evaluations are carried out on benchmark datasets and an E-shaped magnetic core image database, in which all images are captured by the designed AVI system. Experimental results show that the proposed methods achieve an improved performance when comprising with the state-of-the-art methods in this application.
Unsupervised image segmentation is greatly useful in many vision-based applications. In this paper, we aim at the unsupervised low-key image segmentation. In low-key images, dark tone dominates the ...background, and gray level distribution of the foreground is heterogeneous. They widely exist in the areas of space exploration, machine vision, medical imaging,
etc
. In our algorithm, a novel active contour model with the probability density function of gamma distribution is proposed. The flexible gamma distribution gives a better description for both of the foreground and background in low-key images. Besides, an unsupervised curve initialization method is designed, which helps to accelerate the convergence speed of curve evolution. The experimental results demonstrate the effectiveness of the proposed algorithm through comparison with the CV model. Also, one real-world application based on our approach is described in this paper.