Representing multiscale N-folded symmetry in color texture

We develop methods to extract semantically meaningful symmetries from color images. These symmetries are defined within and between color bands using complex moments computed from the output of a bank of orientation and scale selective filters. From this representation, we derive a set of features which are invariant to rotation, scale, and illumination conditions. Experimental results are provided to show the performance of this set of features for classification and image database partitioning.

Dense micro-block difference (DMD) has achieved good performance in gray texture representation and classification. However, its performance is not satisfactory when representing color texture. To alleviate this problem, we propose a novel color texture representation method based on Completed Extremely Nonnegative DMD (CEN-DMD) in this paper. In particular, we first use DMD to model interchannel features and interchannel features of color texture images. Considering that negative value is meaningless in a digital image, we perform a nonnegative operation during the difference process. Due to that the maximum value in a nonnegative difference patch represents a significant difference, we construct the Extremely Nonnegative DMD (EN-DMD) by fusing the maximum values of the intrachannel features and the maximum of interchannel features, and further build Completed Extremely Nonnegative DMD (CEN-DMD) by fusing EN-DMDs at five scales and the global feature of the color texture images. Finally, the Fisher Vector is used to encode the CEN-DMD to obtain a color texture descriptor. Experimental results on five published standard color texture datasets (CUReT, Colored Brodatz, VisTex, USPTex and KTH-TIPS) reveal that CEN-DMD is effective when compared to the thirteen representative color texture classification methods.

A perception-based approach to content-based image representation and retrieval is proposed in this paper. We consider textured images and propose to model their textural content by a set of features having a perceptual meaning and their application to content-based image retrieval. We present a new method to estimate a set of perceptual textural features, namely coarseness, directionality, contrast, and busyness. The proposed computational measures can be based upon two representations: the original images representation and the autocorrelation function (associated with original images) representation. The set of computational measures proposed is applied to content-based image retrieval on a large image data set, the well-known Brodatz database. Experimental results and benchmarking show interesting performance of our approach. First, the correspondence of the proposed computational measures to human judgments is shown using a psychometric method based upon the Spearman rank-correlation coefficient. Second, the application of the proposed computational measures in texture retrieval shows interesting results, especially when using results fusion returned by each of the two representations. Comparison is also given with related works and show excellent performance of our approach compared to related approaches on both sides: correspondence of the proposed computational measures with human judgments as well as the retrieval effectiveness.

Color texture contains a large amount of spectral and spatial structure that can be exploited for recognition. Recent work has demonstrated that spatial filters offer a convenient means of extracting illumination invariant spatial information from a color image. In this paper, we address the problem of deriving optimal fillers for illumination-invariant color texture discrimination. Color textures are represented by a set of illumination-invariant features that characterize the color distribution of a filtered image region. Given a pair of color textures, we derive a spatial filter that maximizes the distance between these textures in feature space. We provide a method for using the pair-wise result to obtain a filter that maximizes discriminability among multiple classes. A set of experiments on a database of deterministic and random color textures obtained under different illumination conditions demonstrates the improved discriminatory power achieved by using an optimized filler.

In the present paper, a set of robust texture features based on morphological residues of opening or closing by reconstruction is introduced. This set of features is proven much more robust to noise than those based on traditional morphological residues. The robustness to noise of this set of features is investigated in detail. An algorithm is constructed to select the most discriminating feature subset. In texture classification experiments, it is found that this feature subset bears high robustness to various noises even when the noise levels are quite high.

The use of classification techniques for machine health monitoring and fault diagnosis has been popular in recent years. System response in form of time series data can be used to identify type of defect, severity of defect etc. However, a central issue with time series classification is that of identifying appropriate features for classification. In this paper, we explore a new feature set based on a delay differential equations (DDEs). DDEs have been used recently for extracting features for classification but have never been used to classify system responses. The Duffing oscillator and Van der Pol–Duffing (VDP-D) oscillator are used as dynamic systems, and the responses are classified into self-similar groups. Responses with the same period should belong to the same group. Misclassification rate is used as an indicator of the efficacy of the feature set. The proposed feature set is compared to a statistical feature set, a power spectral coefficient feature set and a wavelet coefficient feature set. In work described in this paper, a density estimation algorithm called DBSCAN is used as the classification algorithm. The proposed DDE-based feature set is found to be significantly better than the other feature sets for the classifying responses generated by the Duffing system. The wavelet and the power spectral coefficient data sets are not found to be significantly better than the statistical feature set for the Duffing system. None of the feature sets tested are discerning enough on the VDP-D system.

Objectives: To identify a cervical cytology cell image as a representative of the normal or malignant cancerous cell. This work may be used to decide the stage of cervical cancer. Methods and Analysis: Images from the Hervlev university database have been used. Features of nucleus and cytoplasm of cervical cell images are extracted after preprocessing and segmentation. Size, shape and texture features are extracted. Appropriate and adequate features are selected by mining techniques. This feature set is fed to classifiers to decide nature of cells. Based on parameters like true positive rate, true negative rate and precision, a comparative analysis of classifiers is done. The best classifier that clearly discriminates the cells is determined for cervical cancer diagnosis. Findings: Dimensionality reduction of feature set can be done with data mining technique named as Correlation based feature subset selection. Reduced feature set can be used for classification by classifiers like Multilayer perceptron, Bayes classifiers and SVM classifier. Kappa Statistics for binary classifiers is 1 for Multilayer perceptron, 0.9968 for Binary SMO and Bayesnet, 0.9915 for NaiveBayes Classifiers. The same for seven group classifiers is 1 for LibSVM, 0.8535 for Multilayer perceptron, 0.8464 for attribute selected classifier and 0.7286 for Bagging classifier. True positive rate is 1 for Multilayer perceptron, 0.996 for Binary SMO and BayesNet, 0.975 for NaiveBayes classifier. Multilayer perceptron can be used as a classifier in both binary and seven class classifications. Improvement: Existing methods use more features for classification. Here even in the case of seven class classifier, only 11 features are used and desired result has been achieved.

Various applications using a camera system have been developed and deployed commercially to improve our daily life. The performance of camera system is mainly dependent on image quality and illumination conditions. Multiband camera has been developed to provide a wealth of information for image acquisition. In this paper, we developed two applications about image segmentation and face detection using a multiband camera, which is available in four bands consisting of a near infrared and three color bands. We proposed a multiband camera system to utilize two different images i.e. color image extracted from Bayer filter and near infrared images. The experimental results showed the effectiveness of the proposed system.

The aims of this paper are two-fold: to define Gaussian mixture models (GMMs) of colored texture on several feature spaces and to compare the performance of these models in various classification tasks, both with each other and with other models popular in the literature. We construct GMMs over a variety of different color and texture feature spaces, with a view to the retrieval of textured color images from databases. We compare supervised classification results for different choices of color and texture features using the Vistex database, and explore the best set of features and the best GMM configuration for this task. In addition we introduce several methods for combining the ‘color’ and ‘structure’ information in order to improve the classification performances. We then apply the resulting models to the classification of texture databases and to the classification of man-made and natural areas in aerial images. We compare the GMM model with other models in the literature, and show an overall improvement in performance.

Machine learning algorithms are widely used to annotate biological sequences. Most learning algorithms require data to be represented using a vector of features. In the absence of biologically informative features, k-mers generated using the sliding window-based approach are commonly used to represent biological sequences. However, the number of features generated using the sliding window-based approach increases exponentially with the length of the k-mer, k, and as a consequence, the computational time of the learning algorithms also increases. As an alternative to k-mers, we propose to use a community detection approach to identify informative features for sequence classification. The community detection approach is run on a network constructed based on k-mer similarity, and it has the advantage that class labels are not needed when constructing the network. It works by identifying sets of nodes that are densely connected when compared to other nodes of the network. Community detection approaches are computationally expensive, especially for large or densely connected networks. To address this challenge, we propose an efficient strategy that involves running the community detection approach on a set of samples, where each sample is a small subset of the total set of sequences. A set of features is obtained from each sample, and the union of all feature sets is used to represent sequences for classification purposes. Experimental results in both supervised and semi-supervised settings show that the features generated with the community detection approach are more informative than k-mers. Furthermore, we observe that running the community detection approach on a set of small samples of sequences produces better features as compared to running the community detection approach on a set of large samples of sequences.

In this paper, a novel algorithm which integrates the RGB color histogram and texture features for content based image retrieval. A new set of two- dimensional (2-D) M-band dual tree complex wavelet transform (M_band_DT_CWT) and rotated M_band_DT_CWT are designed to improve the texture retrieval performance. Unlike the standard dual tree complex wavelet transform (DT_CWT), which gives a logarithmic frequency resolution, the M-band decomposition gives a mixture of a logarithmic and linear frequency resolution. Most texture image retrieval systems are still incapable of providing retrieval result with high retrieval accuracy and less computational complexity. To address this problem, we propose a novel approach for image retrieval using M_band_DT_CWT and rotated M_band_DT_CWT (M_band_DT_RCWT) by computing the energy, standard deviation and their combination on each subband of the decomposed image. To check the retrieval performance, two texture databases are used. Further, it is mentioned that the databases used are Brodatz gray scale database and MIT VisTex Color database. The retrieval efficiency and accuracy using proposed features is found to be superior to other existing methods. Index Terms—Color; M-band wavelets; Feature Extraction; M-band dual tree complex wavelets; Image Retrieval.

Identification of the short transient waveform, called a spike, in the cortical electroencephalogram (EEG) plays an important role during diagnosis of neurological disorders such as epilepsy. It has been suggested that artificial neural networks (ANN) can be employed for spike detection in the EEG, if suitable features are provided as input to an ANN. In this paper, we explore the performance of neural network-based classifiers using features selected by algorithms suggested by four previous investigators. Of these, three algorithms model the spike by mathematical parameters and use them as features for classification while the fourth algorithm uses raw EEG to train the classifier. The objective of this paper is to examine if there is any inherent advantage to any particular set of features, subject to the condition that the same data are used for all feature selection algorithms. Our results suggest that artificial neural networks trained with features selected using any one of the above three algorithms as well as raw EEG directly fed to the ANN will yield similar results.

In Part I ["The shiftable complex directional pyramid-Part I: Theoretical Aspects," IEEE TRANSACTIONS on Signal Processing, vol. 56, no. 10, October 2008], we proposed a novel image decomposition called the shiftable complex directional pyramid. The shiftable, multiresolution and multidirectional image decomposition is constructed by the pyramidal dual-tree directional filter bank (PDTDFB). In this Part II, we consider the design and implementation issues of the PDTDFB and its applications in image processing. The PDTDFB is composed of a multiresolution filter bank (FB) and a pair of directional filter banks (DFB); the design of each FB is considered separately. The binary tree structure of the DFB is modified to minimize the phase delay factor in order to reduce the border artifact in case symmetric extension is used. The filters presented in this paper are designed in frequency domain, and approximated by finite-impulse-response (FIR) filters. Numerical experiments show that the filters from the dual branches approximate the Hilbert transform of those from the primal ones, and the resulting PDTDFB yields acceptable shiftability. The shiftable pyramid decomposition is then tested in image processing applications, including image denoising, texture retrieval and seismic image migration.

Identification of important image features for pork and turkey ham classification using colour and wavelet texture features and genetic selection

Investigation of acoustic and visual features for pig cough classification