Problem Of Feature Extraction Research Articles

Semisupervised Transfer Component Analysis for Domain Adaptation in Remote Sensing Image Classification

In this paper, we study the problem of feature extraction for knowledge transfer between multiple remotely sensed images in the context of land-cover classification. Several factors such as illumination, atmospheric, and ground conditions cause radiometric differences between images of similar scenes acquired on different geographical areas or over the same scene but at different time instants. Accordingly, a change in the probability distributions of the classes is observed. The purpose of this work is to statistically align in the feature space an image of interest that still has to be classified (the target image) to another image whose ground truth is already available (the source image). Following a specifically designed feature extraction step applied to both images, we show that classifiers trained on the source image can successfully predict the classes of the target image despite the shift that has occurred. In this context, we analyze a recently proposed domain adaptation method aiming at reducing the distance between domains, Transfer Component Analysis, and assess the potential of its unsupervised and semisupervised implementations. In particular, with a dedicated study of its key additional objectives, namely the alignment of the projection with the labels and the preservation of the local data structures, we demonstrate the advantages of Semisupervised Transfer Component Analysis. We compare this approach with other both linear and kernel-based feature extraction techniques. Experiments on multi- and hyperspectral acquisitions show remarkable cross- image classification performances for the considered strategy, thus confirming its suitability when applied to remotely sensed images.

Pose estimation-dependent identification method for field moth images using deep learning architecture

Due to the varieties of moth poses and cluttered background, traditional methods for automated identification of on-trap moths suffer problems of incomplete feature extraction and misidentification. A novel pose estimation-dependent automated identification method using deep learning architecture is proposed in this paper for on-trap field moth sample images. To deal with cluttered background and uneven illumination, two-level automated moth segmentation was created for separating moth sample images from each trap image. Moth pose was then estimated in terms of either top view or side view. Suitable combinations of texture, colour, shape and local features were extracted for further moth description. Finally, the improved pyramidal stacked de-noising auto-encoder (IpSDAE) architecture was proposed to build a deep neural network for moth identification. The experimental results on 762 field moth samples by 10-fold cross-validation achieved a good identification accuracy of 96.9%, and indicated that the deployment of the proposed pose estimation process is effective for automated moth identification.

Filter bank learning for signal classification

This paper addresses the problem of feature extraction for signal classification. It proposes to build features by designing a data-driven filter bank and by pooling the time–frequency representation to provide time-invariant features. For this purpose, our work tackles the problem of jointly learning the filters of a filter bank with a support vector machine. It is shown that, in a restrictive case (but consistent to prevent overfitting), the problem boils down to a multiple kernel learning instance with infinitely many kernels. To solve such a problem, we build upon existing methods and propose an active constraint algorithm able to handle a non-convex combination of an infinite number of kernels. Numerical experiments on both a brain–computer interface dataset and a scene classification problem prove empirically the appeal of our method.

Surface Feature Extraction Based on Curvelet Transform from Point Cloud Data

Aiming at multi directions analysis problem of surface feature extraction from point cloud data, Curvelet transform is introduced to multi directions analysis of point cloud data. Based on the preprocessing of location and expansion, second-generation discrete Curvelet transform is used to analyze point cloud data. Curvelet transform coefficients are processed to enhance the contour of point cloud data. Nonlinear function is used to process Curvelet transform coefficients of coarse layer. Compromise for soft and hard thresholds is used to process Curvelet transform coefficients of detail layer. Piecewise nonlinear function is used to process Curvelet transform coefficients of fine layer. The data point is reconstructed from the enhanced Curvelet transform coefficient with Curvelet inverse transformation. Initial surface feature is achieved with edge detection. The precise surface feature is achieved with morphological dilation and erosion to filter edge without real shape significance. Example of part point cloud data of brake shell shows the proposed surface feature extraction method can accurately extract surface feature from data point.

Analysis and Research on Combination Feature Extraction Method of EEG Singnal

EEG feature extraction problem is studied in this paper. EEG analysis is the core content of the Brain-computer interface technology research. How to effectively extract the reflect people's behavior intention characteristic from EEG signals, it's a hot spot in this neighborhood research. According to the characteristics of EEG signal, the single method of feature extraction can't describe the characteristics of the signal very well. So We have own designed experiment, and put forward a combination feature extraction method, which contains calculation the maximum Lyapunov exponent and use wavelet packet transform to calculate the rhythm average energy with wavelet energy entropy, then, the extract feature vector is inputted into the binary tree support vector machine (SVM) and the extreme learning machine (ELM), respectively. From the recognition result show that, when use the combination method of feature extraction to solve the problem of feature extraction and classification about this subject acquisition EEG, it's feasible and effective. At the same time, it also provides a new thought and method.

Research on Locality Preserving Discriminant Projection Algorithm Based on Gabor for Face Expression Recognition

For the problem of features extraction and dimensionality reduction of expression recognition, the paper proposes Gabor Locality Preserving Discriminant Projection (GLPDP) algorithm, which is based on Gabor Wavelet. Firstly, we use Gabor wavelet transform to have an expression feature extraction. Secondly, we improved the locality preserving projection (LPP) algorithm, introducing scatter difference in the LPP objective function to increase divergence constraints among the sample classes and extracts more discriminated features while having the dimensionality reduction. Finally, we use the nearest neighbor classifier to have a classification for expression category. The effectiveness of the proposed methods is validated through the experimental results on JAFFE and Cohn-Kanade Facial expression databases.

Exploration of Genetic Programming Optimal Parameters for Feature Extraction from Remote Sensed Imagery

Abstract. Evolutionary computation is used for improved information extraction from high-resolution satellite imagery. The utilization of evolutionary computation is based on stochastic selection of input parameters often defined in a trial-and-error approach. However, exploration of optimal input parameters can yield improved candidate solutions while requiring reduced computation resources. In this study, the design and implementation of a system that investigates the optimal input parameters was researched in the problem of feature extraction from remotely sensed imagery. The two primary assessment criteria were the highest fitness value and the overall computational time. The parameters explored include the population size and the percentage and order of mutation and crossover. The proposed system has two major subsystems; (i) data preparation: the generation of random candidate solutions; and (ii) data processing: evolutionary process based on genetic programming, which is used to spectrally distinguish the features of interest from the remaining image background of remote sensed imagery. The results demonstrate that the optimal generation number is around 1500, the optimal percentage of mutation and crossover ranges from 35% to 40% and 5% to 0%, respectively. Based on our findings the sequence that yielded better results was mutation over crossover. These findings are conducive to improving the efficacy of utilizing genetic programming for feature extraction from remotely sensed imagery.

METHODS AND CHARACTERISTICS OF LOCALITY-PRESERVING TRANSFORMATIONS IN THE PROBLEMS OF COMPUTATIONAL INTELLIGENCE

The problem of the development of mathematical support for data dimensionality reduction is solved. Its results can be used to automate the process of diagnostic and recognizing model construction by precedents. The set of rapid transformations from the original multidimensional space to the one-dimensional axis was firstly proposed. They provide a solution of the feature extraction and feature selection problems. The complex of indicators characterizing the properties of transformations was firstly proposed. On the basis of the proposed indicators the set of criteria was defined. It facilitate comparison and selection of the best transformations and results of their work in diagnosis and recognition problems solving on the basis of computational intelligence methods. The software realizing proposed transformations and indicators characterizing their properties was developed. The experimental study of proposed transformations and indicators was conducted, which results allow to recommend the proposed transformations for use in practice.

L1-norm kernel discriminant analysis via Bayes error bound optimization for robust feature extraction.

A novel discriminant analysis criterion is derived in this paper under the theoretical framework of Bayes optimality. In contrast to the conventional Fisher's discriminant criterion, the major novelty of the proposed one is the use of L1 norm rather than L2 norm, which makes it less sensitive to the outliers. With the L1-norm discriminant criterion, we propose a new linear discriminant analysis (L1-LDA) method for linear feature extraction problem. To solve the L1-LDA optimization problem, we propose an efficient iterative algorithm, in which a novel surrogate convex function is introduced such that the optimization problem in each iteration is to simply solve a convex programming problem and a close-form solution is guaranteed to this problem. Moreover, we also generalize the L1-LDA method to deal with the nonlinear robust feature extraction problems via the use of kernel trick, and hereafter proposed the L1-norm kernel discriminant analysis (L1-KDA) method. Extensive experiments on simulated and real data sets are conducted to evaluate the effectiveness of the proposed method in comparing with the state-of-the-art methods.

Features extraction: problem formulation and solving technique

The application of the fractal dynamic analysis method to tensometrical data is results in some variables appearance. There is a problem of features extraction in the paper. Features calculation methods, classification methods and features extraction criteria are discussed. The review of features extraction algorithms are represented.

Recognition Model Based Feature Extraction and Kernel Extreme Learning Machine for High Dimensional Data

High dimensional data such as mass-spectrometric and near-infrared spectrum are always used in disease diagnosis and product quality monitoring. Aim at the nonlinear feature extraction and low learning speed problems, a novel modeling approach combined principal component analysis (PCA) with kernel extreme learning machine (KELM) is proposed. The extracted features using PCA algorithms are fed into nonlinear classification based KELM with fast learning speed. The numbers of the features are selected according the classification performance. The experimental results based on the mass-spectrometric data in the benchmark demonstrate that the proposed approach has better performance. This approach can also be used to target recognition based on radar data.

Multiple-instance discriminant analysis

Multiple-instance discriminant analysis (MIDA) is proposed to cope with the feature extraction problem in multiple-instance learning. Similar to MidLABS, MIDA is also derived from linear discriminant analysis (LDA), and both algorithms can be treated as multiple-instance extensions of LDA. Different from MidLABS which learns from the bag level, MIDA is designed from the instance level. MIDA consists of two versions, i.e., binary-class MIDA (B-MIDA) and multi-class MIDA (M-MIDA), which are utilized to cope with binary-class (standard) and multi-class multiple-instance learning tasks, respectively. The block coordinate ascent approach, by which we seek positive prototypes (the most positive instance in a positive bag is termed as the positive prototype of this bag) and projection vectors alternatively and iteratively, is proposed to optimize B-MIDA and M-MIDA to obtain lower dimensional transformation subspaces. Extensive experiments empirically demonstrate the effectiveness of B-MIDA and M-MIDA in extracting discriminative components and weakening class-label ambiguities for instances in positive bags.

An Improved Abnormal Behavior Feature Detection Algorithm of Network Information Based on Fractional Fourier Transform

The feature extraction and detection problem of network information abnormal behavior was researched in this paper. The network attack tended to ambiguity in hidden, and the abnormal behavior of network information is referred as a data signal series, and it was existed in the network information space with strong interference. Traditional detection method was hard to detect the abnormal signal. On the basis of fractional Fourier transform (FRFT), an improved abnormal behavior feature detection algorithm of network information was proposed. The properties such as energy gathering and noise suppression of 4-order cumulant slice were taken in advantage. In the post processing of fractional Fourier transform detection, the post processing operator of 4 order cumulant was introduced in the detection algorithm, the post energy was gathered in the fractional Fourier domain, the signal accumulation was likely to be increased, and the interference noise could be restrained effectively. Simulation results show that the improved algorithm has perfect noise suppression performance, and it can detect and extract the abnormal behavior feature in the network space to maximum. The detection performance is improved greatly, and the research result can be applied in the network information warfare and network security areas.

大规模互联网图像检索与模式挖掘

The explosive growth of web images not only brings many technical challenges to image search, but also provides almost unlimited training data and new ideas to various computer vision problems. This paper presents a brief historical review of three stages of image retrieval, with a particular emphasis on the impact of large-scale web images to image retrieval. Based on the review, the paper discusses the fundamental problem of feature extraction in image retrieval, and the recent research trend on visual pattern mining to bridge the semantic gap. According to their representation granularity, the paper divides visual patterns into ve categories and introduces their related work respectively, which also shows the great importance of big data to visual pattern mining.

A Retrieval Method of Ancient Chinese Character Images

Global and local image retrieval of ancient Chinese characters is a helpful means for character research work. Because of the characteristics of ancient Chinese characters such as complex structures and variant shapes, there exist many theoretical and technical problems in feature extraction, clustering and retrieval of these images. A retrieval method was established with the strategy of “clustering before matching” for the ancient Chinese character images by integrating the structural features of them. Firstly, preprocessed character image area was divided into elastic meshes and the directional elements were extracted to form the feature vectors. Then, K-means algorithm was employed to cluster the character images within global and local areas. Finally, the similar images within selected areas were searched in corresponding cluster and the obtained images were provided to users. The experimental results show that this method is helpful for the improvement of the efficiency of ancient Chinese character study.

Mapping of Rescue Environment Based on NDT Scan Matching

This paper studied the mapping problem for rescue robots based on laser scan matching and extend Kalman filtering (EKF). Because of the non-structural rescue environments, it is hard to extract typical features. Scan matching method based on normal distribution transform (NDT) can avoid the hard feature extraction problem by estimation of the probability distribution of laser scan data. By fusing NDT scan matching with EKF framework, the NDT-EKF SLAM algorithm was proposed, which can effectively and precisely build maps for rescue environment. Experiment results show that NDT-EKF SLAM algorithm is more precise than algorithms based solely on scan-matching.

Welding Defect Classification of Ultrasonic Detection Based on PCA and KNN

Aiming to problems of welding defect classification in the ultrasonic detection, according to the characteristics of welding defect, a classification method based on PCA and KNN is proposed in order to solve the problem of ultrasonic testing signal feature extraction and defect recognition. For the impact of the redundant attributes of feature extraction on the classification accuracy, in this paper, an ultrasonic flaw feature extraction algorithm based on PCA is proposed. Ultrasonic flaw intelligent classification is always a difficult problem in NDT. KNN algorithm is proposed to classify the different defects. Compared with BP algorithm, experiment results show that the model based on PCA and KNN can get stable classification results, high accuracy, and can effectively improve the classification efficiency.

Image Recognition Technology Based on Artificial Intelligence

In respect of the classification of current image retrieval technology and the existing issues, the paper put forward a method designed for image semantic feature extraction based on artificial intelligence. The new method has solved the tough problem of image semantic feature extraction, by fusing fuzzy logic, genetic algorithm and artificial neural network altogether, which greatly improved the efficiency and accuracy of image retrieval.

Training Data Reduction and Classification Based on Greedy Kernel Principal Component Analysis and Fuzzy C-Means Algorithm

Nonlinear feature extraction used standard Kernel Principal Component Analysis (KPCA) method has large memories and high computational complexity in large datasets. A Greedy Kernel Principal Component Analysis (GKPCA) method is applied to reduce training data and deal with the nonlinear feature extraction problem for training data of large data in classification. First, a subset, which approximates to the original training data, is selected from the full training data using the greedy technique of the GKPCA method. Then, the feature extraction model is trained by the subset instead of the full training data. Finally, FCM algorithm classifies feature extraction data of the GKPCA, KPCA and PCA methods, respectively. The simulation results indicate that the feature extraction performance of both the GKPCA, and KPCA methods outperform the PCA method. In addition of retaining the performance of the KPCA method, the GKPCA method reduces computational complexity due to the reduced training set in classification.

A review of object representation based on local features

Object representation based on local features is a topical subject in the domain of image understanding and computer vision. We discuss the defects of global features in present methods and the advantages of local features in object recognition, and briefly explore state-of-the-art recognition methods using local features, especially the main approaches of local feature extraction and object representation. To clearly explain these methods, the problem of local feature extraction is divided into feature region detection, feature region description, and feature space optimization. The main components and merits of these steps are presented. Technologies for object presentation are classified into three types: vector space, sliding window, and structure relationship models. Future development trends are discussed briefly.