Image Recognition Applications Research Articles

WE‐DE‐207B‐02: Detection of Masses On Mammograms Using Deep Convolutional Neural Network: A Feasibility Study

Purpose:To develop a deep convolutional neural network (DCNN)‐based computer‐aided diagnosis (CAD) system for detecting the masses in digital mammographic images.Methods:A DCNN architecture, which consists of 5 convolutional layers and 3 fully connected layers, is constructed in this study. The DCNN parameters are then trained by the following two procedures. We first train the DCNN using about 1.3 million natural images for classification of 1,000 categories. Then, we modify the last fully connected layer and subsequently train the modified DCNN using 1,656 mammographic region of interest (ROI) images for two categories classification: mass and normal.Results:The trained DCNN is tested by using 198 mammographic ROI images including 99 mass images and 99 normal images. The experimental results show that the sensitivity of the mass detection is about 89.9% and the false positive is 19.2%. These results demonstrated that the DCNN has a potential for mammographic CAD.Conclusion:In recent years, the DCNN, as one of the most successful techniques in deep learning technology, made a remarkable impact on image recognition application. For medical image recognition, however, its performance is uncertainty because collecting a large amount of training image data for a particularly medical image modality is difficult. In this study, our preliminary experiments demonstrated a feasibility to apply the DCNN in mammographic CAD system. To the best of our knowledge, this study is also the first demonstration of DCNN for detecting the masses in mammographic images.

Real-time hardware\u2013software embedded vision system for ITS smart camera implemented in Zynq SoC

The article demonstrates the usefulness of heterogeneous System on Chip (SoC) devices in smart cameras used in intelligent transportation systems (ITS). In a compact, energy efficient system the following exemplary algorithms were implemented: vehicle queue length estimation, vehicle detection, vehicle counting and speed estimation (using multiple virtual detection lines), as well as vehicle type (local binary features and SVM classifier) and colour (k-means classifier and YCbCr colourspace analysis) recognition. The solution exploits the hardware–software architecture, i.e. the combination of reconfigurable resources and the efficient ARM processor. Most of the modules were implemented in hardware, using Verilog HDL, taking full advantage of the possible parallelization and pipeline, which allowed to obtain real-time image processing. The ARM processor is responsible for executing some parts of the algorithm, i.e. high-level image processing and analysis, as well as for communication with the external systems (e.g. traffic lights controllers). The demonstrated results indicate that modern SoC systems are a very interesting platform for advanced ITS systems and other advanced embedded image processing, analysis and recognition applications.

Rolling Bearing Fault Diagnosis Based on STFT-Deep Learning and Sound Signals

The main challenge of fault diagnosis lies in finding good fault features. A deep learning network has the ability to automatically learn good characteristics from input data in an unsupervised fashion, and its unique layer-wise pretraining and fine-tuning using the backpropagation strategy can solve the difficulties of training deep multilayer networks. Stacked sparse autoencoders or other deep architectures have shown excellent performance in speech recognition, face recognition, text classification, image recognition, and other application domains. Thus far, however, there have been very few research studies on deep learning in fault diagnosis. In this paper, a new rolling bearing fault diagnosis method that is based on short-time Fourier transform and stacked sparse autoencoder is first proposed; this method analyzes sound signals. After spectrograms are obtained by short-time Fourier transform, stacked sparse autoencoder is employed to automatically extract the fault features, and softmax regression is adopted as the method for classifying the fault modes. The proposed method, when applied to sound signals that are obtained from a rolling bearing test rig, is compared with empirical mode decomposition, Teager energy operator, and stacked sparse autoencoder when using vibration signals to verify the performance and effectiveness of the proposed method.

Learning and intelligence can happen everywhere, a case study: learning via Non-uniform 1D rulers with applications in image classification and recognition

In this paper, we presented a non-uniform 1D ruler model and applied it in various image classification and image recognition scenarios, and some are for military technology usage. Our model is very simple, elegant and original, which is solved by convex quadratic programming. It has wide applications in pattern recognition and intelligent multimedia data analysis. We believe that a new research topic, namely, numeric calibration, has started, which is parallel to dimensionality reduction, feature selection, or metric learning etc. Our methods can be used as a pre-processing step for metric learning methods, in which, our learned calibrated feature space is used as input for them. The various combinations of our methods and metric learning methods, may lead to new interesting research problems.

Large-scale image recognition based on parallel kernel supervised and semi-supervised subspace learning

Kernel discriminant subspace learning technique is effective to exploit the structure of image dataset in the high-dimensional nonlinear space. However, for large-scale image recognition applications, this technique usually suffers from large computational burden. Although some kernel accelerating methods have been presented, how to greatly reduce computing time and simultaneously keep favorable recognition accuracy is still challenging. In this paper, we introduce the idea of parallel computing into kernel subspace learning and build a parallel kernel discriminant subspace learning framework. In this framework, we firstly design a random non-overlapping equal data division strategy to divide the whole training set into several subsets and assign each computational node a subset. Then, we separately learn kernel discriminant subspaces from these subsets without mutual communications and finally select the most appropriate subspace to classify test samples. Under the built framework, we propose two novel kernel subspace learning approaches, i.e., parallel kernel discriminant analysis (PKDA) and parallel kernel semi-supervised discriminant analysis (PKSDA). We show the superiority of the proposed approaches in terms of time complexity as compared with related methods, and provide the fundamental supports for our framework. For experiment, we establish a parallel computing environment and employ three public large-scale image databases as experiment data. Experimental results demonstrate the efficiency and effectiveness of the proposed approaches.

Better than SIFT?

Independent evaluation of the performance of feature descriptors is an important part of the process of developing better computer vision systems. In this paper, we compare the performance of several state-of-the art image descriptors including several recent binary descriptors. We test the descriptors on an image recognition application and a feature matching application. Our study includes several recently proposed methods and, despite claims to the contrary, we find that SIFT is still the most accurate performer in both application settings. We also find that general purpose binary descriptors are not ideal for image recognition applications but perform adequately in a feature matching application.

Adaptive Convolutional Neural Network and Its Application in Face Recognition

Convolutional neural network (CNN) has more and more applications in image recognition. However, the structure of CNN is often determined after a performance comparison among the CNNs with different structures, which impedes the further development of CNN. In this paper, an adaptive convolutional neural network (ACNN) is proposed, which can determine the structure of CNN without performance comparison. The final structure of ACNN is determined by automatic expansion according to performance requirement. First, the network is initialized by a one-branch structure. The system average error and recognition rate of the training samples are set to control the expansion of the structure of CNN. That is to say, the network is extended by global expansion until the system average error meets the requirement and when the system average error is satisfied, the local network is expanded until the recognition rate meets the requirement. Finally, the structure of CNN is determined automatically. Besides, the incremental learning for new samples can be achieved by adding new branches while keeping the original network unchanged. The experiment results of face recognition on ORL face database show that there is a better tradeoff between the consumption of training time and the recognition rate in ACNN.

Similarity Measure Based on Distance of Dual Hesitant Fuzzy Sets and Its Application in Image Feature Comparison and Recognition

In this paper, we propose the distance-based similarity measure of dual hesitant fuzzy sets (DHFSs) and apply it to the image feature comparison and recognition. Based on distance-based similarity measure of DHFS, we establish a comprehensive assessment method of image features to compare the images, where five criteria are represented by DHFSs to assess the image features. By the proposed method, we can determine ranking order of the six alternative images. The assessment results show that the proposed method is simple and effective to solve the problem of image feature compari- son and image recognition, and provides a new assessment method for computer image processing experts.

ПОВЫШЕНИЕ ПРОИЗВОДИТЕЛЬНОСТИ ФРАКТАЛЬНОГО КОДИРОВАНИЯ ИЗОБРАЖЕНИЙ С ПОМОЩЬЮ ПЕРЦЕПТИВНЫХ ХЕШ-ФУНКЦИЙ

The application of fractal encodingfor image recognition is considered, system requirements to computing resources of classic fractal encoding system implementation is analyzed. The proposal for improving performance characteristics is suggested and implemented, obtained results and comparison of two implementations are provided in the last part of article.

Local Directional Texture Pattern image descriptor

Deriving an effective image representation is a critical step for a successful automatic image recognition application. In this paper, we propose a new feature descriptor named Local Directional Texture Pattern (LDTP) that is versatile, as it allows us to distinguish person’s expressions, and different landscapes scenes. In detail, we compute the LDTP feature, at each pixel, by extracting the principal directions of the local neighborhood, and coding the intensity differences on these directions. Consequently, we represent each image as a distribution of LDTP codes. The mixture of structural and contrast information makes our descriptor robust against illumination changes and noise. We also use Principal Component Analysis to reduce the dimension of the multilevel feature set, and test the results on this new descriptor as well.

Content‐based image retrieval over IEEE 802.11b noisy wireless networks

SummaryMobile devices such as smartphones and tablets are widely used in everyday life to perform a variety of operations, such as e‐mail exchange, connection to social media, bank/financial transactions, and so on. Moreover, because of the large growth of multimedia applications, video and image transferring and sharing via a wireless network is becoming increasingly popular. Several modern mobile applications perform information retrieval and image recognition. For example, Google Goggles is an image recognition application that is used for searches based on pictures taken by handheld devices. In most of the cases, image recognition procedure is an image retrieval procedure. The captured images or a low‐level description of them are uploaded online, and the system recognizes their content by retrieving visually similar pictures. Taking into account the last comment, our goal in this paper is to evaluate the process of image retrieval/recognition over an Institute of Electrical and Electronics Engineers 802.11b network, operating at 2.4 GHz. Our evaluation is performed through a simulated network configuration, which consists of a number of mobile nodes communicating with an access point. Throughout our simulations, we examine the impact of several factors, such as the existence of a strong line of sight during the communication between wireless devices. Strong line of sight depends on the fading model used for the simulations and has an effect on BER. We have used a large number of image descriptors and a variety of scenarios, reported in the relative literature, in order to comprehensively evaluate our system. To reinforce our results, experiments were conducted on two well‐known images databases by using 10 descriptors from the literature. Copyright © 2014 John Wiley & Sons, Ltd.

基于变分偏微分方程的红外图像增强算法研究

Infrared images typically have the problems of low-contrast and blur edges.This article introduces the variation theory of partial differential equations to enhance the infrared images.By reconstructing image gradient field distribution,detailed texture and low-contrast can be improved.The processing of noise consists the cut of amplitude in the transform of gradient field.Based on experiment results,the original image gray gradient can be increased by 2~3times and therefore significantly enhances the faint information and alleviates the edge effects.This approach can effectively overcome the problem of losing similar gray scale distribution caused by the histogram class algorithms.Therefore,it gives qualified image information for image recognition and tracking application.

The Application of Image Recognition in Mobile Solar Power Device: Based on Light Intensity Detection in Shaded Area

In the comparative analysis of image characteristics of shadow-free and shaded area, an algorithm for shaded area, avoiding the area whose illumination is weak and tending to the area with strong light via looking for the strongest and weakest light area, is established, which can thereby improve the efficiency of power generation. The result demonstrates that the method can find the region with the strongest sunlight fast and accurately.

The Application of Image Recognition Based on BP Neural Networks in Automatic Steel Rolling

A BP neural networks is presented for billet character recognition. Firstly, by a series of image processing techniques, the character’feature in the billet character region of the video image gathered by frame grabber is abstracted. Secondly, the BP neural networks algorithm is employed for character recognition. Application results show that the image recognition based BP neural networks can performs well in billet character recognition, and the method presented is speedy, efficient and of high value in practice.

Full-range in-plane rotation measurement for image recognition with hybrid digital-optical correlator

A hybrid digital-optical correlator (HDOC) based on volume holographic memory is able to compute the correlation of images at a high speed. HDOC is suitable for real-time image processing and has potential usage in big data processing areas. A 7500-channel HDOC system is experimentally set up, and the target image is correlated with all the channels. The large number of parallel correlation channels could contribute to the precise rotation measurement as well as the translation measurement. In the image recognition applications, the target image involves rotation distortion with respect to the template images. A method with two coarse-fine steps is proposed to measure the rotation at a full range of 360 deg. In the coarse step, the target image is rotated 36 times at an increment of 10 deg. The 36 new images are sent into the HDOC to compute with the template images. Each new image corresponds to a correlation matrix. By searching the smallest value throughout the 36 minimums of the 36 correlation matrixes, the rotation of the target image is narrowed into ±5 deg . In the fine step, the new image is rotated another 10 times at an increment of 1 deg. The rotation measurement error is <0.3 deg .

OPTIMAL ALGORITHM FOR PREDICTING STUDENTS ACADEMIC PERFORMANCE

Machine learning has been successfully applied to numerous domains such as pattern recognition, image recognition, fraud detection, medical diagnosis, banking, bioinformatics, commodity trading, computer games and various control applications. Recently, this paradigm is been employed to enhance and evaluate higher education tasks. The focus of this work is on identifying the optimal algorithm suitable for predicting first-year tertiary students academic performance based on their family background factors and previous academic achievement. One thousand five hundred (1,500) enrolment records of students admitted into computer science programme Babcock University, Nigeria between 2001 and 2010 was used. The students’ first year academic performance was measured by Cumulative Grade Point Average (CGPA) at the end of the first session and the previous academic achievement was measured by SSCE grade score and UME score. Waikato Environment for Knowledge Analysis (WEKA) was used to generate 10 classification models( five decision tree algorithms -Random forest, Random tree, J48, Decision stump and REPTree and five rule induction algorithms –JRip, OneR, ZeroR, PART, and Decision table)  and a multilayer perceptron, an artificial neural network function. These algorithms were compared using 10-fold cross validation and hold-out method considering accuracy level, confusion matrices and CPU time to determine the optimal model. This work will be taken further by designing a framework of predictive system based on the rules generated from the optimal model.

A new distance measure based on generalized Image Normalized Cross-Correlation for robust video tracking and image recognition

We propose two novel distance measures, normalized between 0 and 1, and based on Normalized Cross-Correlation for image matching. These distance measures explicitly utilize the fact that for natural images there is a high correlation between spatially close pixels. Image matching is used in various computer vision tasks, and the requirements to the distance measure are application dependent. Image recognition applications require more shift and rotation robust measures. In contrast, registration and tracking applications require better localization and noise tolerance. In this paper, we explore different advantages of our distance measures, and compare them to other popular measures, including Normalized Cross-Correlation (NCC) and Image Euclidean Distance (IMED). We show which of the proposed measures is more appropriate for tracking, and which is appropriate for image recognition tasks.

Image Recognition and Force Measurement Application in the Humanoid Robot Imitation

This paper studies and implements motion imitation interaction between a humanoid robot and a human. To achieve the implementation, three main processes are required. The first is the human motion data acquisition; the second is the motion data modification; and the last is the ankle angle adjustment on the supporting foot (or feet) of the humanoid robot. In the human motion data acquisition, we use a webcam to recognize the 13 marks that are pasted on the human's body and then calculate and record the relative positions of the marks for each motion into the motion database. In the motion data modification, the recorded motion data are modified by computer simulation to guarantee that the zero moment point of the robot is inside the stable region. Finally, based on the force sensor measurement on the soles of the robot, the ankle angles of the supporting foot (or feet) are adjusted such that the humanoid robot is capable of imitating the human motion with a balance status in real time. The experimental results demonstrate that the humanoid robot successfully imitates a series of human gymnastic motions.

DESIGNING A SOPC FOR FACE RECOGNITION USING WMPCA ALGORITHM

A flexible accelerator hardware for full-search vector quantization (VQ) has been developed as a component for a system on a programmable chip (SoPC) to use in real-time image compression and recognition applications. Nowadays, FPGA and its SoPC (System on Programmable Chip) tools are powerful enough to efficiently develop a flexible hardware accelerator for VQ application. In addition, one of statistical analysis methods, weighted modular principal component analysis, has showed efficiencies in recognition applications. In this paper, a parallel architecture for online face recognition using weighted modular principal component analysis (WMPCA) and its system-on-programmable-chip (SoPC) implementation are discussed.

Statistical Computations on Grassmann and Stiefel Manifolds for Image and Video-Based Recognition

In this paper, we examine image and video-based recognition applications where the underlying models have a special structure—the linear subspace structure. We discuss how commonly used parametric models for videos and image sets can be described using the unified framework of Grassmann and Stiefel manifolds. We first show that the parameters of linear dynamic models are finite-dimensional linear subspaces of appropriate dimensions. Unordered image sets as samples from a finite-dimensional linear subspace naturally fall under this framework. We show that an inference over subspaces can be naturally cast as an inference problem on the Grassmann manifold. To perform recognition using subspace-based models, we need tools from the Riemannian geometry of the Grassmann manifold. This involves a study of the geometric properties of the space, appropriate definitions of Riemannian metrics, and definition of geodesics. Further, we derive statistical modeling of inter and intraclass variations that respect the geometry of the space. We apply techniques such as intrinsic and extrinsic statistics to enable maximum-likelihood classification. We also provide algorithms for unsupervised clustering derived from the geometry of the manifold. Finally, we demonstrate the improved performance of these methods in a wide variety of vision applications such as activity recognition, video-based face recognition, object recognition from image sets, and activity-based video clustering.