Image identification system based on an optical broadcast neural network and a pulse coupled neural network preprocessor stage

We describe the concept of a vision system based on an optoelectronic hardware neural processor. The proposed system is composed of a pulse coupled neural network (PCNN) preprocessor stage that converts an input image into a temporal pulsed pattern. These pulses are inputs to the optical broadcast neural network (OBNN) processor, which classifies the input pattern between a set of reference patterns based on a pattern matching strategy. The PCNN is to provide immunity to the scale, rotation, and translation of objects in the image. The OBNN provides high parallelism and a high speed hardware neural processor.     

Image identification system based on an optical broadcast neural network processor

We describe the implementation of a vision system based on a hardware neural processor. The architecture of the neural network processor has been designed to exploit the computational characteristics of electronics and the communication characteristics of optics in an optimal manner, thus it is based on an optical broadcast of input signals to a dense array of processing elements. The vision system has been built by use of a prototype implementation of a neural network processor with discrete optic and optoelectronic devices. It has been adapted to work as a Hamming classifier of the images taken with a 128 x 128 complementary metal-oxide semiconductor image sensor. Its results, performance characteristics of the image classification system, and an analysis of its scalability in size and speed, with the improvement of the optoelectronic neural processor, are presented.     

Damage detection of CFRP laminates using electrical resistance measurement and neural network

As carbon fibers are electrical conductors, the measurement of the electrical resistance appears to be a valuable technique for the in situ detection of various types of damage in carbon fiber reinforced polymers (CFRP) laminates. In such cases, carbon fibers are both the reinforcement and the sensor to detect damage in CFRP laminates. The damage-detecting method of CFRP laminates by electrical resistance measurement that are investigated in this study is made possible by attaching electrodes on the surface of the CFRP structures without special manufacturing.

In this paper, we investigate the electrical resistance change as a damage parameter of fatigue damage such as the degradation of residual strength and stiffness. The measured stiffness and electrical resistance change during fatigue tests showed a very similar trend of change. This is because cumulative fatigue damage is represented by the degradation of residual stiffness; these damages also cause change in electrical resistance. Thus, we can use this change in electrical resistance as a damage parameter. We also predict the future damage of composite laminates in fatigue loading from electrical resistance damage model by following a stiffness degradation model. Electrical resistance gradually increased as the stiffness reduced, and showed a very abrupt change when final fatigue failure was imminent. The predicted value showed good agreement with the experimental data except in the final stage, where stiffness and electrical resistance changed abruptly.     

Color reduction using local features and a kohonen self-organized feature map neural network

This paper proposes a new method for reducing the number of colors in an image. The proposed approach uses both the image color components and local image characteristics to feed a Kohonen self-organized feature map (SOFM) neural network. After training, the neurons of the output competition layer define the proper color classes. The final image has the dominant image colors and its texture approaches the image local characteristics used. To speed up the entire algorithm and reduce memory requirements, a fractal scanning subsampling technique can be used. The method is applicable to all types of color images and can be easily extended to accommodate any type of spatial characteristics. Several experimental and comparative results are presented. © 1999 John Wiley & Sons, Inc. Int J Imaging Syst Technol 10, 404–409, 1999     

A combination of rough-based feature selection and RBF neural network for classification using gene expression data

This paper presents a novel rough-based feature selection method for gene expression data analysis. It can find the relevant features without requiring the number of clusters to be known a priori and identify the centers that approximate to the correct ones. In this paper, we attempt to introduce a prediction scheme that combines the rough-based feature selection method with radial basis function neural network. For further consider the effect of different feature selection methods and classifiers on this prediction process, we use the NaIve Bayes and linear support vector machine as classifiers, and compare the performance with other feature selection methods, including information gain and principle component analysis. We demonstrate the performance by several published datasets and the results show that our proposed method can achieve high classification accuracy rate.     

基于神经网络的在线计算机图像识别系统

以二值型自适应共振理论（ART－1）神经网络为识别核心设计了一个应用于生产流水线的计算机识别系统，它可以对生产线上的零件和产品的文字和符号进行实时识别，作自动记录。该系统具有学习和识别速度快、识别率高（＞96％），可以灵活改变识别对象，应用范围广等特点。     

Breast cancer diagnosis from mammographic images using optimized feature selection and neural network architecture

Breast cancer is one of the deadly diseases in women that have raised the mortality rate of women. An accurate and early detection of breast cancer using mammogram images is still a complex task. Hence, this article proposes a novel breast cancer detection model, which included five major phases: (a) preprocessing, (b) segmentation, (c) feature extraction, (d) feature selection, and (e) classification. The input mammogram image is initially preprocessed using contrast limited adaptive histogram equalization (CLAHE) and median filtering. The preprocessed image is then subjected to segmentation via the region growing algorithm. Subsequently, geometric features, texture features and gradient features are extracted from the segmented image. Since the length of the feature vector is large, it is essential to select the optimal features. Here, the selection of optimal features is done by a hybrid optimization algorithm. Once the optimal features are selected, they are subjected to the classification process involving the neural network (NN) classifier. As a novelty, the weight of NN is selected optimally to enhance the accuracy of diagnosis (benign and malignant). The optimal feature selection as well as the weight optimization of NN is accomplished by merging the Lion algorithm (LA) and particle swarm optimization (PSO), named as velocity updated lion algorithm (VU‐LA). Finally, a performance‐based evaluation is carried out between VU‐LA and the existing models like, whale optimization algorithm (WOA), gray wolf optimization (GWO), firefly (FF), PSO, and LA.     

Experimental results for improving the matrix condition using a hybrid optical system

We present preliminary experimental results for implementing the "blurred trajectories" method on three parallel optics (PO) systems. The "main" system and "auxiliary" optics were simple laboratory graded lenses attached to an iris diaphragm. When applying the blurred trajectories method we first show an improvement in the matrix condition, as the matrix condition number decreased in a range of factors of 3 to 418 relative to the main system. Following that, image restoration by weak regularization was performed so that the system matrix condition dominated the restoration process. It was shown that the restoration results of the PO are better than those of the main system and the auxiliary optics separately. In addition, the quality of the restoration follows the system's matrix condition. The improvement in the matrix condition achieved by the PO system improved the immunity to detection noise. Finally, a comparison to Wiener filtering restoration shows that it is also generally inferior to the proposed method.     

基于均值特征和改进深度神经网络的说话人识别算法

为提高神经网络在说话人识别应用中的识别性能,提出基于高斯增值矩阵特征和改进深度卷积神经网络的说话人识别算法.算法首先通过最大后验概率提取基于梅尔频率倒谱系数(Mel Frequency Cepstrum Coefficient,MFCC)特征的高斯均值矩阵,并对特征进行噪声适应性补偿,以增强信号的帧间关联和说话人特征信...     

基于模糊神经网络的智能故障诊断专家系统

发动机是车辆的核心部件,及时有效地发现并排除故障,对降低维修费用,减少经济损失,增加发动机工作时的可靠性,避免事故发生具有重大的意义。以某型号发动机为研究对象,运用测试技术、信号处理、小波分析、神经网络和模糊控制理论,提出了基于模糊神经网络的智能故障诊断系统。建立了发动机故障信号采集试验台,在试验台上人工模拟3种转速下6种工况,通过加速度传感器采集正常工况和异常工况的振动信号,之后利用小波包技术进行消噪处理,并提取出故障信号的特征值,作为网络训练和测试的样本数据。用样本数据训练和检测自适应模糊神经网络,完成对信号的离线模式识别,之后以测试样本数据实现在线故障诊断,通过仿真分析,取得了很好的诊断效果。与传统的BP神经网络故障诊断方法进行对比,无论在诊断精度上还是学习速度上,模糊神经网络在故障诊断中更具有优势。同时,在专家系统的理论基础上,将模糊神经网络与专家系统进行信息融合,实现数据接口通信,利用网络的自学习能力建立智能故障诊断数据库和诊断规则库,通过程序语言快速高效的设计出智能诊断系统。最后,通过发动机故障诊断实例仿真分析,验证了基于模糊神经网络的智能故障诊断专家系统的可行性。     

An integrated approach for process monitoring using wavelet analysis and competitive neural network

A novel framework involving both a detection module and a classification module is proposed for the recognition of the six main types of process signals. In particular, a multi-scale wavelet filter is used for denoising and its performance is compared with that of single-scale linear filters. Moreover, two kinds of competitive neural networks, based on learning vector quantization (LVQ) and adaptive resonance theory (ART), are adopted for the task of pattern classification and benchmarking. Our results show that denoising through a wavelet filter is best for pattern classification, and the classification accuracy with respect to six predefined categories using a LVQ-X network is a little better than using an ART network. However, when an unexpected novel pattern occurs within the process, LVQ will force the novel pattern to be classified into one of those predefined categories that is most similar to the novel pattern. On the contrary, ART will automatically construct a new class when the similarity measured between the novel pattern and the most similar category is too small to be incorporated. Therefore, under the consideration of the stability–plasticity dilemma, our simplified ART network based on multi-scale wavelet denoising provides a more promising way to adapt unexpected novel patterns.     

Technique for detecting flaws in metallic surfaces using an optical system with phase-type blazed gratings

In this paper, we evaluate a novel flaw-detection technique for metallic surfaces based on the use of phase-type blazed gratings. Transparent blazed gratings were prepared by the soft nanoimprint method involving the transfer of a template (a reflective grating structure used for spectroscopy) onto silicone rubber. The blazed gratings were then integrated into an imaging system to observe the reflective metal sample. Due to the low-pass-filtering properties of the gratings, the captured image was notably blurred. This characteristic aids in flaw detection on metallic surfaces because the captured image is adequate to distinguish flaws in the targeted area on the basis of the texture of the rough surface, including any other structures that were unintended. The use of double-sided gratings with crossing grating vectors was found to be efficient for homogenous low-pass filtering. Such flaw-detection techniques are expected to be useful for conducting quality inspections of rolled steel plates since the surface contains both a rough surface and undesirable flaws.     

采用BP神经网络优化的振动信号压缩感知方法

无线传感网络逐渐应用于结构健康监测,但是因能耗问题难以实现长期、高频的数据采集工作。压缩感知技术可利用少量的采样点重构原始信号,有望降低无线传感网络的能耗。实测振动信号因受到噪声干扰而导致稀疏性有限,常用于压缩感知的LASSO算法难以精确求解稀疏系数,进而影响振动信号重构效果。引入BP神经网络优化LASSO算法解得的稀疏系数,BP神经网络经ADAM优化算法训练后,可有效提升振动信号重构精度。用三层框架结构的模拟加速度数据和广州塔的监测加速度数据验证方法的有效性,并探讨了正则化参数和优化迭代次数的影响。结果表明,基于BP神经网络优化的压缩感知方法的信号重构效果在不同压缩率下均优于非优化的压缩感知方法。     

A chemical synthesized Al-doped PbS nanoparticles hybrid composite for optical and electrical response

Structural, morphological, optical and electrical investigations of pure and Al-doped lead sulfide (PbS) nanoparticles hybrid composite was synthesized by simple chemical route. The detail analysis of the nanoparticle morphology of hybrid composites through optical investigation, phase purity and crystalline size had been characterized by using X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscope (TEM), ultraviolet spectroscopy (UV), photoluminescence (PL). The lower angle XRD results confirmed that the phase purity and average crystalline size of the pure and Al doped PbS nanoparticles were determined by using the Debye–Scherrer’s formula. The average grain sizes of the pure and the Al-doped PbS nanoparticles were calculated and found to be 22 and 16 nm respectively. Surface morphology analysis was carried out by using SEM and TEM analysis. The surface morphology of pure and Al doped PbS nanoparticles is without any pinholes or cracks and hence they appear to be densely packed with spherical shaped grains. The optical properties of pure and Al-doped PbS analyzed using UV–Vis. absorption spectroscopy and Photoluminiscence (PL) spectra. The band gap values for the pure and the Al-doped PbS nanoparticles were found to be 1.94 and 2.04 eV respectively. The dielectric properties of the Al-doped PbS nanoparticle composites typical response e.g. dielectric constant, dielectric loss, and AC conductivity were analyzed at various frequencies and temperatures.     

An optical technique for detecting defects in thin plates and diaphragms

A non-destructive test is proposed for thin plates and diaphragm elements that consists of recording the images of a coarse grating reflected by the plates or elements when they are laterally loaded. A defective region is indicated by an obvious distortion of the lines constituting the grating image.     

An artificial neural network for modeling reliability, availability and maintainability of a repairable system

The paper explores the application of artificial neural networks to model the behaviour of a complex, repairable system. A composite measure of reliability, availability and maintainability parameters has been proposed for measuring the system performance. The artificial neural network has been trained using past data of a helicopter transportation facility. It is used to simulate behaviour of the facility under various constraints. The insights obtained from results of simulation are useful in formulating strategies for optimal operation of the system.     

An intelligent control for linear ultrasonic motor using interval type-2 fuzzy neural network

An interval type-2 fuzzy neural network (IT2FNN) is developed for the position control of a thetas-axis motion-control stage using a linear ultrasonic motor to confront the uncertainties of the motion-control stage. A T2FNN consists of a type-2 fuzzy linguistic process as the antecedent part and a three-layer interval neural network as the consequent part. A general T2FNN is computationally intensive due to the complexity of reducing type 2 to type 1. Therefore an IT2FNN is adopted to simplify the computational process. Moreover, the developed IT2FNN combines the merits of an interval type-2 fuzzy logic system and a neural network. Furthermore, the parameter-learning of the IT2FNN, which is based on the supervised gradient decent method using a delta adaptation law, is performed on line. Experimental results show that the dynamic behaviours of the proposed IT2FNN control system are more effective and robust with regard to uncertainties than the type-1 FNN control system.     

A hybrid convolutional neural network model to detect COVID-19 and pneumonia using chest X-ray images

A hybrid convolutional neural network (CNN)-based model is proposed in the article for accurate detection of COVID-19, pneumonia, and normal patients using chest X-ray images. The input images are first pre-processed to tackle problems associated with the formation of the dataset from different sources, image quality issues, and imbalances in the dataset. The literature suggests that several abnormalities can be found with limited medical image datasets by using transfer learning. Hence, various pre-trained CNN models: VGG-19, InceptionV3, MobileNetV2, and DenseNet are adopted in the present work. Finally, with the help of these models, four hybrid models: VID (VGG-19, Inception, and DenseNet), VMI(VGG-19, MobileNet, and Inception), VMD (VGG-19, MobileNet, and DenseNet), and IMD(Inception, MobileNet, and DenseNet) are proposed. The model outcome is also tested using five-fold cross-validation. The best-performing hybrid model is the VMD model with an overall testing accuracy of 97.3%. Thus, a new hybrid model architecture is presented in the work that combines three individual base CNN models in a parallel configuration to counterbalance the shortcomings of individual models. The experimentation result reveals that the proposed hybrid model outperforms most of the previously suggested models. This model can also be used in the identification of diseases, especially in rural areas where limited laboratory facilities are available.