Search

The supervised dictionary learning methods have made considerable achievements in the field of pattern recognition. In order to make the learned coefficients have intraclass similarity and interclass incoherence, many dictionary learning methods have been proposed using different structured constraints. However, these constraints often have the disadvantages of complexity, time-consuming, and poor interpretability. More importantly, the existing dictionary learning often ignores the inherent structure of the label matrix. In this paper, we propose a dictionary pair learning based on the structured classifier and perform classifier learning and structured coefficients learning simultaneously. Our main idea is to use the extended label matrix and the invertible constraints on the classification transformation matrix to utilize the structure of the label matrix to impose structured constraints on coefficients. Moreover, the l 21 -norm is added to force the analysis dictionary to focus on more important features. Experimental results on multiple databases prove that our proposed method has better classification performance than several state-of-the-art dictionary learning methods. [ABSTRACT FROM AUTHOR]

Keyword spotting techniques are becoming cost-effective solutions for information retrieval in handwritten documents. We explore the extension of the single-word, line-level probabilistic indexing approach described in our previous works to allow for page-level search of queries consisting in Boolean combinations of several single-keywords. We propose heuristic rules to combine the single-word relevance probabilities into probabilistically consistent confidence scores of the multi-word boolean combinations. An empirical study, also presented in this paper, evaluates the search performance of word-pair queries involving AND and OR Boolean operations. Results of this study support the proposed approach and clearly show its effectiveness. Finally, a web-based demonstration system based on the proposed methods is presented. [ABSTRACT FROM AUTHOR]

Segmentation of dried amnion biomaterial tends to produce invalid (outlier) contour point detections due to texture and colour inhomogeneity of the biomaterial. In this paper, a novel implementation of a non-stationary Bayesian learning process for outlier contour point removal of amnion segmentations is presented. This outlier removal method is independent to algorithms used for the contour detection. The Bayesian process uses a non-stationary kernel to learn a function with complex shape that maps image features in a region-of-interest around each contour point to a discrete output. Based on this output, a contour point can be determined as valid or invalid (outlier). The hyper-parameters of the non-stationary kernel are learned by maximising the marginal likelihood of the combined likelihood of data and the prior of the kernel parameters. Moreover, a novel combination of gradient-ascend and harmonic heuristic search methods is presented to find the optimal hyper-parameters. To validate the method, experiments are conducted to detect and ignore invalid contour points on amnion biomaterial images. A comparison of the proposed method with a logistic regression classification as the baseline is performed. The results show that the proposed method can significantly improve the contour detection by removing outliers and, hence, can reduce waste of uncut biomaterials. [ABSTRACT FROM AUTHOR]

This paper presents an optical character/text recognition (OCR) system for cursive scripts like those of Arabic, Urdu, Persian, Kurdish, etc. This OCR system is a large-scale one in the sense of architecture, training data size, and state-of-the-art performance. The paper introduces the theoretical derivation and experimental assessment of our two main contributions deployed in this OCR. The first contribution is the design of a new autonomously normalized, lossless, and compact feature vector that enables the production of a truly robust omnifont OCR system for cursive scripts with an ASR-like HMM-based architecture. Half of the components in this feature vector are analogs and the other half are discrete, which obstructs the use of continuous Gaussian mixtures to model an aggregate of such features and mandates the use of discrete HMMs instead, which in turn necessitates the deployment of vector clustering and quantization. The second contribution is a new soft (i.e., probabilistic) vector quantization (VQ) scheme, as opposed to conventional hard-deciding VQ that we analytically derive and then deploy to alleviate overfitting and boost the robustness of our OCR against different kinds of obtrusive variances. We present experimental evidence of these benefits of the presented soft VQ scheme to our OCR. Other machine-learning systems with VQ modules may also deploy soft VQ to obtain the same benefits. [ABSTRACT FROM AUTHOR]

Control chart pattern analysis is a crucial task in statistical process control. There are various types of nonrandom patterns that may appear on the control chart indicating that the process is out of control. The presence of nonrandom patterns manifests that a process is affected by assignable causes, and corrective actions should be taken. From a process control point of view, identification of nonrandom patterns can provide clues to the set of possible causes that must be searched; hence, the troubleshooting time could be reduced in length. In this paper, we discuss two implementation modes of control chart pattern recognition and introduce a new research issue concerning pattern displacement problem in the process of control chart analysis. This paper presents a neural network-based pattern recognizer with selected features as inputs. We propose a novel application of statistical correlation analysis for feature extraction purposes. Unlike previous studies, the proposed features are developed by taking the pattern displacement into account. The superior performance of the feature-based recognizer over the raw data-based one is demonstrated using synthetic pattern data. [ABSTRACT FROM AUTHOR]

Dimensionality reduction plays a major role in face recognition. Discriminant analysis (DA) and principal component analysis (PCA) are two of the most important approaches in this field. In particular, subclass discriminant analysis (SDA) is a well-known scheme for feature extraction and dimensionality reduction. It is widely used in many high-dimensional data-driven applications, namely face recognition and image retrieval. It is also found to be applicable under various scenarios. However, it has high cost in time and space given the need for an eigendecomposition involving the scatter matrices, known as the singularity problem. This limitation is caused by the high-dimensional space of data, particularly when dimensions exceed the number of observations. Recent advances widely reported that 2D methods with matrix-based representation perform better than the traditional 1D vector-based ones. In this paper, we propose a novel 2D-SDA algorithm to avoid the "curse of dimensionality" and address the singularity issue. The performance of the proposed algorithm is evaluated for face recognition in terms of recognition performance and computational cost. Experiments are conducted on four benchmark face databases and compared to several competitive 1D and 2D methods based on PCA and DA. Results show that 2DSVD achieves the best recognition performance at low dimensions. In particular, 2D-SDA works significantly better on large-sized data sets where intra-class variation is the most important. [ABSTRACT FROM AUTHOR]

The Euler number of a binary image is an important topological feature for many image processing, image analysis, pattern recognition, and computer vision applications. This paper proposes a new run-based Euler number computation algorithm. The conventional run-based algorithm processes rows of the given image one-by-one from top to bottom in a single phase. For each row, it finds the runs in the row and records the start and end locations of each run to compute neighbor runs. In contrast, our algorithm calculates the Euler number of an image in two phases. In the first phase, we process odd rows alternately to find runs and only record its end location. In the second phase, we process each of the remaining even rows to find runs and calculate neighboring runs between the current row and the rows immediately above and below using the recorded run data. Using this method, the number of accesses required to compute the Euler number decreases in almost all cases. Analysis of the time complexity and experimental results demonstrate that our algorithm outperforms conventional Euler number computation algorithms. [ABSTRACT FROM AUTHOR]

Intuitionistic fuzzy sets, introduced by Atanassov, offer a new possibility to describe in a more adequate way many real problems. An important tool for determining the degree of similarity between two objects is represented by similarity measures. Different types of similarity measures for intuitionistic fuzzy sets were proposed and used in various applications. More existing similarity measures involve little or no the hesitancy degree in intuitionistic fuzzy sets. The capacity of a similarity measure is determined by the form of expression and the information contained therein; more information used implies a greater power to distinguish. In this paper, we propose a family of intuitionistic fuzzy similarity measures that contain the hesitancy degree in their expression. This family represents a generalization of some known measures. For particular values of parameters, various similarity measures can be obtained: some of them are known but others are new, as will be shown in sections "The proposed intuitionistic fuzzy similarity measures" and "Conclusion". In order to evaluate the performance of these measures, our experimental study is oriented in two basic directions: one shows how much are reasonable the proposed measures, and other describes the possibilities of application in solving some real-world problems. The experiments prove that our measures give the same results as the measures used as comparison. [ABSTRACT FROM AUTHOR]

Microalgae counting is used to measure biomass quantity. Usually, it is performed in a manual way using a Neubauer chamber and expert criterion, with the risk of a high error rate. Scenedesmus algae can build coenobia consisting of 1, 2, 4 and 8 cells. The amount of algae of each coenobium helps to determine the amount of lipids, proteins, and other substances in a given sample of a algae crop. The knowledge of the quantity of those elements improves the quality of bioprocess applications. This paper addresses the methodology for automatic identification of Scenedesmus microalgae (used in the methane production and food industry) and applies it to images captured by a digital microscope. The use of contrast adaptive histogram equalization for pre-processing, and active contours for segmentation are presented. The calculation of statistical features (histogram of oriented gradients, Hu and Zernike moments) with texture features (Haralick and local binary patterns descriptors) are proposed for algae characterization. Classification of coenobia achieves accuracies of 98.63% and 97.32% with support vector machine and artificial neural network, respectively. According to the results, it is possible to consider the proposed methodology as an alternative to the traditional technique for algae counting. In addition, the database used for the developing of the proposed methodology is publicly available. [ABSTRACT FROM AUTHOR]

Orientation-based coding approaches have recently been widely employed for face and palmprint recognition where generally, one starts with a set of Gabor filters to extract orientation information and then proceeds to code dominant orientations as features for each point of the palmprint. However, as the Gabor filter is developed to model two-dimensional receptive fields of simple cells in straits cortex, it might not be our best choice when dealing with curved and complex structures inherent in the palmprint texture. Motivated by this intuition, this paper shows that Gabor filters are a subset of a bigger family of filters which we refer to as generalized Gabor filter (GGF). Depending on the values of its parameters, a GGF takes a rather diverse shapes and orientations, which results in a potentially finer feature extraction capability. We show this improved capability by employing GGFs in the palmprint verification process. In applying our method, two different sub-banks of GGFs are defined for the orientation-based feature extraction of palmprints, and when compared with Gabor filters, it will be shown that GGFs have the upper hand in capturing orientation features. Furthermore, compared with the competitive code—one of the well-known orientation-based coding methods—the number of employed orientations is reduced to half. This would automatically compensate for a double usage of the filter banks, which otherwise could increase the time complexity of using GGFs. These ideas are further elaborated using a set of experiments on PolyU II and PolyU 2D/3D palmprint databases. The results show the preeminence of using GGFs both in terms of accuracy and efficiency. [ABSTRACT FROM AUTHOR]

The use of computer-readable visual codes became common in our everyday life both in industrial environments and for private use. The reading process of visual codes consists of two steps, namely, localization and data decoding. In this paper we examine the localization step of visual codes using conventional and deep rectifier neural networks. They are also evaluated in the discrete cosine transform domain and shown to be efficient, which makes full decompression unnecessary for setups involving JPEG images. This approach is also efficient from a storage viewpoint and computation cost viewpoint, since camera hardware can provide a JPEG stream as output in many cases. The use of neural networks implemented on graphics processing unit allows real-time automatic code object localization. In our earlier studies, the proposed approach was evaluated on the most popular code type, quick response code, and some other 2D codes as well. Here, we also prove that deep rectifier networks are also suitable for 1D barcode localization and present extensive evaluation and comparison to state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Motion sensing plays an important role in the study of human movements, motivated by a wide range of applications in different fields, such as sports, health care, daily activity, action recognition for surveillance, assisted living and the entertainment industry. In this paper, we describe how to classify a set of human movements comprising daily activities using a wearable motion capture suit, denoted as FatoXtract. A probabilistic integration of different classifiers recently proposed is employed herein, considering several spatiotemporal features, in order to classify daily activities. The classification model relies on the computed confidence belief from base classifiers, combining multiple likelihoods from three different classifiers, namely Naïve Bayes, artificial neural networks and support vector machines, into a single form, by assigning weights from an uncertainty measure to counterbalance the posterior probability. In order to attain an improved performance on the overall classification accuracy, multiple features in time domain (e.g., velocity) and frequency domain (e.g., fast Fourier transform), combined with geometrical features (joint rotations), were considered. A dataset from five daily activities performed by six participants was acquired using FatoXtract. The dataset provided in this work was designed to be extremely challenging since there are high intra-class variations, the duration of the action clips varies dramatically, and some of the actions are quite similar (e.g., brushing teeth and waving, or walking and step). Reported results, in terms of both precision and recall, remained around 85 %, showing that the proposed framework is able to successfully classify different human activities. [ABSTRACT FROM AUTHOR]

Several feature-based pattern recognition algorithms have been developed during the past decade. These algorithms rely on identifying keypoints in an image and assigning a descriptor to each point based on the composition of their surrounding region. Comparison of the descriptors of keypoints found in two images enables these algorithms to match similar objects within those images. The dependence of these algorithms' performance on the similarity of the internal structure of objects makes them susceptible to modifications that change this internal structure. In this paper, we first compare the relative performance of some major feature-based algorithms in finding similar objects surrounded by geometrical noise. Next, we add several noise and transformation types to target objects and re-evaluate the performance of these algorithms under the resulting structural changes. Our results provide insights on the relative strengths of these algorithms in the presence and absence of several noise and transformation types. In addition, these findings allow us to identify modification types that can better inhibit the performance of these algorithms. The resulting insight can be used in applications that need to build resistance against such algorithms, e.g., in developing CAPTCHAs that need to be resistant to recognition attacks. [ABSTRACT FROM AUTHOR]

The dangerous, overgrown vegetation/trees under high voltage transmission lines right-of-ways (ROWs) have caused severe blackouts/flashovers due to interference with power lines which leads to short circuiting among the conductors. Therefore, these dangerous encroachments are monitored periodically along the electrical distribution networks ROWs through visual inspection, or by airborne system. Each of these methods has its own attributes and limitations and have proved to be costly, time consuming and not much accurate. In these circumstances, it is necessary for the electrical utilities to review their vegetation management practices so as to avoid incidents of unintended encroachments. In tropical countries, overgrown vegetation is a common cause for power line failure. This paper proposes an innovative concept of utilizing a single camera for monitoring dangerous vegetation (trees, shrubs and plants, etc.) under transmission lines ROWs. The main focus is on using an imaging device (camera) integrated on each transmission pole to automate inspection for the vegetation encroachments endangering the transmission lines. These cameras are envisioned to be connected wirelessly to each other, forming a series of wireless camera networks that can be monitored remotely. A single camera mounted on power poles acquires images and sends them wirelessly to the base station. At base station, algorithm (software) trained by image processing and pattern recognition techniques is used to identify (height, depth, and width of encroached vegetation, etc.) excess vegetation encroachments within and outside ROWs. The performance evaluation of a real time developed test-bed scenario proves the feasibilities of integrating the method for transmission line maintenance. [ABSTRACT FROM AUTHOR]

Anomaly detection in time-series data is a relevant problem in many fields such as stochastic data analysis, quality assurance, and predictive modeling. Markov models are an effective tool for time-series data analysis. Previous approaches utilizing Markov models incorporate transition matrices (TMs) at varying dimensionalities and resolutions. Other analysis methods treat TMs as vectors for comparison using search algorithms such as the nearest neighbors comparison algorithm, or use TMs to calculate the probability of discrete subsets of time-series data. We propose an analysis method that treats the elements of a TM as random variables, parameterizing them hierarchically. This approach creates a metric for determining the 'normalcy' of a TM generated from a subset of time-series data. The advantages of this novel approach are discussed in terms of computational efficiency, accuracy of anomaly detection, and robustness when analyzing sparse data. Unlike previous approaches, this algorithm is developed with the expectation of sparse TMs. Accounting for this sparseness significantly improves the detection accuracy of the proposed method. Detection rates in a variety of time-series data types range from (97 % TPR, 2.1 % FPR) to (100 % TPR, <0.1 % FPR) with very small sample sizes (20-40 samples) in data with sparse transition probability matrices. [ABSTRACT FROM AUTHOR]

Moment invariants are an especially important research topic in pattern recognition. There are many kinds of moment invariants published in the literature already. However, there is a need to further improve them, especially under the noisy environments. In this article, we develop a new set of moment invariants by means of ridgelet function. The ridgelet function is capable of capturing line features in an image, which is a particularly important property in pattern recognition. It is well-known that every curve can be approximated by short line segments, so ridgelet moment invariants should be good at robust pattern recognition. We can prove that this set of moments is invariant to the rotation of 2D images. Experiments show that our proposed ridgelet moment invariants are better than the Gaussian–Hermite moments, the Fourier–wavelet descriptor, and Zernike's moment invariants for one Chinese character database and one 2D shape database. Furthermore, our proposed ridgelet moment invariants can do an excellent job for noise-robust pattern recognition. [ABSTRACT FROM AUTHOR]

Imbalanced classification has been a major challenge for machine learning because many standard classifiers mainly focus on balanced datasets and tend to have biased results toward the majority class. We modify entropy fuzzy support vector machine (EFSVM) and introduce instance-based entropy fuzzy support vector machine (IEFSVM). Both EFSVM and IEFSVM use the entropy information of k-nearest neighbors to determine the fuzzy membership value for each sample which prioritizes the importance of each sample. IEFSVM considers the diversity of entropy patterns for each sample when increasing the size of neighbors, k, while EFSVM uses single entropy information of the fixed size of neighbors for all samples. By varying k, we can reflect the component change of sample's neighbors from near to far distance in the determination of fuzzy value membership. Numerical experiments on 35 public and 12 real-world imbalanced datasets are performed to validate IEFSVM, and area under the receiver operating characteristic curve (AUC) is used to compare its performance with other SVMs and machine learning methods. IEFSVM shows a much higher AUC value for datasets with high imbalance ratio, implying that IEFSVM is effective in dealing with the class imbalance problem. [ABSTRACT FROM AUTHOR]

Sign language is the most important means of communication for deaf people. Given the lack of familiarity of non-deaf people with the language of deaf people, designing a translator system which facilitates the communication of deaf people with the surrounding environment seems to be necessary. The system of translating the sign language into spoken languages should be able to identify the gestures in sign language videos. Consequently, this study provides a system based on machine vision to recognize the signs in continuous Persian sign language video. This system generally consists of two main phases of sign words extraction and their classification. Several stages, including tracking and separating the sign words, are conducted in the sign word extraction phase. The most challenging part of this process is separation of sign words from video sequences. To do this, a new algorithm is presented which is capable of detecting accurate boundaries of words in the Persian sign language video. This algorithm decomposes sign language video into the sign words using motion and hand shape features, leading to more favorable results compared to the other methods presented in the literature. In the classification phase, separated words are classified and recognized using hidden Markov model and hybrid KNN-DTW algorithm, respectively. Due to the lack of proper database on Persian sign language, the authors prepared a database including several sentences and words performed by three signers. Simulation of proposed words boundary detection and classification algorithms on the above database led to the promising results. The results indicated an average rate of 93.73 % for accurate words boundary detection algorithm and the average rate of 92.4 and 92.3 % for words recognition using hands motion and shape features, respectively. [ABSTRACT FROM AUTHOR]

In this study, we first examine entropy and similarity measure of Atanassov's intuitionistic fuzzy sets, and define a new entropy. Meanwhile, a construction approach to get the similarity measure of Atanassov's intuitionistic fuzzy sets is introduced, which is based on entropy. Since the independence of elements in a set is usually violated, it is not suitable to aggregate the values for patterns by additive measures. Based on the given entropy and similarity measure, we study their application to Atanassov's intuitionistic fuzzy pattern recognition problems under fuzzy measures, where the interactions between features are considered. To overall reflect the interactive characteristics between them, we define three Shapley-weighted similarity measures. Furthermore, if the information about the weights of features is incompletely known, models for the optimal fuzzy measure on feature set are established. Moreover, an approach to pattern recognition under Atanassov's intuitionistic fuzzy environment is developed. [ABSTRACT FROM AUTHOR]

Dry eye syndrome is a prevalent disease which affects a wide range of the population and has a negative impact on their daily activities, such as driving or working with computers. Its diagnosis and monitoring require a battery of tests which measure different physiological characteristics. One of these clinical tests consists in capturing the appearance of the tear film using the Doane interferometer. Once acquired, the interferometry images are classified into one of the five categories considered in this research. The variability in appearance makes the use of a computer-based analysis system highly desirable. For this reason, a general methodology for the automatic analysis and categorization of interferometry images is proposed. The development of this methodology included a deep study based on several techniques for image texture analysis, three color spaces and different machine learning algorithms. The adequacy of this methodology was demonstrated, achieving classification rates over 93 %. Also, it provides unbiased results and allows important time savings for experts. [ABSTRACT FROM AUTHOR]