for Journals by Title or ISSN for Articles by Keywords help
 Subjects -> ELECTRONICS (Total: 179 journals)
Similar Journals
 Sensing and Imaging : An International JournalJournal Prestige (SJR): 0.255 Citation Impact (citeScore): 1Number of Followers: 2      Hybrid journal (It can contain Open Access articles) ISSN (Print) 1557-2072 - ISSN (Online) 1557-2064 Published by Springer-Verlag  [2352 journals]
• K-Means Clustering Optimizing Deep Stacked Sparse Autoencoder
• Authors: Yandong Bi; Peng Wang; Xuchao Guo; Zhijun Wang; Shuhan Cheng
Abstract: Because of the large structure and long training time, the development cycle of the common depth model is prolonged. How to speed up training is a problem deserving of study. In order to accelerate training, K-means clustering optimizing deep stacked sparse autoencoder (K-means sparse SAE) is presented in this paper. First, the input features are divided into K small subsets by K-means clustering, then each subset is input into corresponding autoencoder model for training, which only has fewer nodes in the hidden layer than traditional models. After training, each autoencoder’s trained weights and biases is merged to obtain the next layer’s input features by feedforward network. The above steps are repeated till the softmax layer, then fine-tuning is carried out. Using MNIST-Rotation datasets to train the network that has three hidden layers and each layer has 800 nodes, the improved model has higher classification accuracy and shorter training time when K = 10. With K increasing, the training time is reduced to almost the same as the fine-tuning time but the recognition ability is descended. Compared with the recently stacked denoising sparse autoencoder, the recognition accuracy is improved by 1%, not only the noise factor is not selected but also the training speed is significantly increased. The trained filters from the improved model is also used to train convolutional autoencoder, and it performs better than traditional models. We find that pre-training stage doesn’t need large samples simultaneously, and small samples parallel training reduces the probability of falling into the local minimum.
PubDate: 2019-02-10
DOI: 10.1007/s11220-019-0227-1
Issue No: Vol. 20, No. 1 (2019)

• Pointwise Multi-resolution Feature Descriptor for Spectral Segmentation
• Authors: JingMao Zhang; YanXia Shen
Abstract: In the process of spectral segmentation, it is crucial to compute a reliable affinity matrix with different features of an image. In this paper, we present a method of constructing the affinity matrix based on multi-resolution features extracted from the original features. A pointwise multi-resolution feature descriptor (PMFD) is designed based on spectral graph wavelets, which characterize the topology of the image centered at different pixels. After choosing the scales of interest in our descriptor, a new affinity matrix is constructed based on the extracted features. For large-size affinity matrixes, it is difficult to compute the proposed PMFD for all pixels of an image. Therefore, an approximate algorithm is proposed to compute the PMFD. To demonstrate the effectiveness of our method, a series of experiments on the Berkeley image segmentation dataset are implemented using the PMFD-based spectral segmentation algorithm. A comparison with other image segmentation techniques demonstrates that our method offers significantly improved pointwise spectral segmentation performance.
PubDate: 2019-01-25
DOI: 10.1007/s11220-019-0226-2
Issue No: Vol. 20, No. 1 (2019)

• Multiple Proposals for Continuous Arabic Sign Language Recognition
• Authors: Mohamed Hassan; Khaled Assaleh; Tamer Shanableh
Abstract: The deaf community relies on sign language as the primary means of communication. For the millions of people around the world who suffer from hearing loss, interaction with hearing people is quite difficult. The main objective of sign language recognition (SLR) is the development of automatic SLR systems to facilitate communication with the deaf community. Arabic SLR (ArSLR) specifically did not receive much attention until recent years. This work presents a comprehensive comparison between two different recognition techniques for continuous ArSLR, namely a Modified k-Nearest Neighbor which is suitable for sequential data and Hidden Markov Models (HMMs) techniques based on two different toolkits. Additionally, in this work, two new ArSL datasets composed of 40 Arabic sentences are collected using Polhemus G4 motion tracker and a camera. An existing glove-based dataset is employed in this work as well. The three datasets are made publicly available to the research community. The advantages and disadvantages of each data acquisition approach and classification technique are discussed in this paper. In the experimental results section, it is shown that classification accuracy for sign sentences acquired using a motion tracker are very similar the classification accuracy for sentences acquired using sensor gloves. The modified KNN solution is inferior to HMMs in terms of the computational time required for classification.
PubDate: 2019-01-17
DOI: 10.1007/s11220-019-0225-3
Issue No: Vol. 20, No. 1 (2019)

• Recognition of Weld Penetration During K-TIG Welding Based on Acoustic and
Visual Sensing
• Authors: Tao Zhu; Yonghua Shi; Shuwan Cui; Yanxin Cui
Abstract: In the field of welding process control, on-line monitoring of welding quality based on multi-sensor information fusion has attracted more attention. In order to recognize the penetration state of the Keyhole mode Tungsten Inert Gas welded joint in real time, an acoustic and visual sensing system was established in this paper. The acoustic and visual features that characterize the penetration state of the welded joints in 34 dimensions were extracted and the variation of the acoustic signal and the keyhole geometry were analyzed. In addition, the weighted scoring criterion based on the Fisher distance and the maximum information coefficient (Fisher–MIC) and Support Vector Machine (SVM) model based on cross-validation (CV) are designed as the feature selection method. The feature selection method can evaluate the penetration recognition accuracy of different feature subsets. The experiment results show that the maximum recognition accuracy was 97.1655%, which was performed by the 10-dimension optimal feature subset and the CV–SVM model with particle swarm optimization (PSO–CV–SVM). It is proved that the selected acoustic and visual features can well characterize the penetration state of the welded joints, and the feature selection method and PSO–CV–SVM model have superior performance.
PubDate: 2019-01-02
DOI: 10.1007/s11220-018-0224-9
Issue No: Vol. 20, No. 1 (2019)

• Extended Super Resolution of Hyperspectral Images via Non-negative Sparse
Coding
• Abstract: Often high resolution (HR) RGB images generated by sparse sampling of the visible spectrum fail to produce differentiable modality for computer vision tasks, Hence computer vision tasks have to rely on gross structures in an image like corners, edges etc. instead of just the recorded reflectance of objects or materials at each pixel in a scene. In contrast to RGB, hyperspectral imaging allows pixels to record reflectance of the scene over multiple contiguous bands, which results in rich differentiable modalities. However, hyperspectral imaging, despite having a growing number of applications from agriculture, surveillance, mineralogy, food processing to eye care, is hitherto restricted to low spatial resolution imaging due to sensor hardware limitations. In this paper, we propose a hyperspectral super resolution technique to produce a high resolution (HR) hyperspectral image with a spectral support of 400–1020 nm from a low resolution (LR) hyperspectral image of the same spectral support and a high resolution multispectral (RGB) image with reduced spectral support of 400–700 nm. In the first step, we generate a HR hyperspectral prior by estimating HR hyperspectral band images in the spectral support of 400–700 nm by detail transfer and alternating iterative minimization. In the next step, we use the generated HR prior to further estimate the HR hyperspectral images for 710–1020 nm bands by learning a non-negative dictionary of reflectance spectra signatures of all the materials present in the scene from the LR hyperspectral image with spectral support in 400–1020 nm. With the estimated HR hyperspectral prior and learned dictionary, we predict the non-negative sparse codes for HR hyperspectral band images in the band of 710–1020 nm.
PubDate: 2019-04-17

• Quality Evaluation of RGB Images Reconstructed by Means of Photoacoustic
Signals
• Abstract: Recent researches have demonstrated the usefulness of photoacoustics in non-destructive control, in particular, in the monitoring and diagnosis of works of art. Indeed, it is fundamental to preserve the artworks’ integrity by using techniques not involving direct contact or damaging radiation, or pre-treatments. On the other hand, a lot of artistic heritage consists of paintings that are complex systems, where, often, the presence of highly scattering and semi-opaque materials make useless optical techniques. Consequently, in this context photoacoustics represent a powerful tool. This work is aimed to evaluate the quality of reconstructed RGB images of simple test objects examined by means of photoacoustic signals, in order to confirm the potentiality of this promising investigation method. Only a single-wavelength excitation source at 1064 nm was available and so, it has been necessary to perform some preliminary processings on the sample color images. The original images have been decomposed in R, G and B components; each of them has been converted into grayscale code, printed on transparency film and then investigated through photoacoustics. After that, the three generated photoacoustic images have been recombined to produce the reconstructed RGB image. A complete experimental system has been set to analyse dedicated test objects. The resulting images have been compared to the original ones, by using standard image quality parameters. Similar results are expected to be obtained by using three sources of distinct wavelengths (Red, Green, Blue), making the method easier to apply.
PubDate: 2019-04-15

• Projection Data Smoothing for Low-Dose CT Based on $$\ell _p$$ ℓ p
Regularization
• Abstract: Projection data smoothing is a traditional technique for low-dose computed tomography. The projection data can be modeled as a piecewise smooth function. It’s well known that $$\ell _1$$ regularization of the image gradients tries to recover piecewise constant functions, while $$\ell _2$$ regularization recovers smooth functions. This motivates us to propose the $$\ell _p$$ regularization with $$1<p<2$$ for low-dose projection data smoothing. Besides, the non-stationary Gaussian noise model for the projection data is built into the regularization term. The resulting model is then linearized such that the fast split-Bregman algorithm can be applied. Experiments on simulated projection data as well as real data show that $$\ell _p$$ regularization with $$1<p<2$$ could achieve better reconstruction compared to $$\ell _{1}$$ regularization.
PubDate: 2019-04-11

• An Approach for Increasing Sensitivity of a Tunable Micro Electro
Mechanical Sensor Using Electrostatic Hopping Voltage
• Abstract: The impetus of the study is to present a novel micro electro mechanical system based tunable gyroscope with the possibility of sensitivity enhancement using appropriate electrostatic hopping voltages. The proposed model is a silicon-based clamped–clamped micro beam sandwiched with two piezoelectric layers throughout the entire length. The nonlinear electrostatic forces are applied to the micro beam along its sense and drive mode directions (either lateral sides). The drive mode actuation is a combination of a direct current (DC) and an alternating current voltage; whereas the sense mode actuation is a pure DC voltage. The micro beam oscillates along the drive mode due to the harmonic drive mode excitation; as the micro beam undergoes base rotation, the Coriolis force induces another motion in the direction of the sense mode which is perpendicular to the drive mode direction. The more is the amplitude of the base rotation, the more is the sense mode amplitude. The sense mode amplitude is directly attributed to the magnitude of base rotation. The piezoelectric layers are actuated by a DC voltage which leads to an axial force proportional to the applied DC voltage. Exciting the piezoelectric layers changes the overall stiffness of the micro beam and as a result the operating frequency of the gyroscope becomes tunable. The partial differential equation of the motion is derived using Hamiltonian principle and discretized into two nonlinear ordinary differential equations along the drive and sense mode directions. The shooting method is used to capture the periodic motion orbits and accordingly the frequency response curves. By using Floquet theory the stability of the periodic orbits is determined. Due to the nonlinearity of the governing equations in the vicinity of the primary resonance, the gyroscope exhibits multi-response solution; Applying appropriate hopping voltages, the micro beam is pushed into the attraction basin of the response with higher amplitude and accordingly the sensitivity of the gyroscope is enhanced. The proposed gyroscope not only has the capability of having improved sensitivity but also its operating frequency can be tuned both in forward and backward directions by means of applying appropriate piezoelectric voltage with an appropriate polarity.
PubDate: 2019-04-03

• Performance of the Eye-Safe LRS and Color CCD Camera Under Aerosol
Environments
• Abstract: In this paper, we performed an experiment to compare the observation performance of color CCD camera and the distance measurement characteristics of the eye-safe laser range scanner in a non-visible environment, which simulated atmospheric environments in the reactor building when a severe accident occurred at the nuclear power plant. For the comparison experiments, we fabricated a fog room. Fog particles are sprayed into the fog room until they reach a certain concentration at the time of the severe accident. In this paper, different methods are used to evaluate the performance of the camera and the eye-safe laser range scanner. To verify the observation performance of the color CCD camera under the dense aerosol environments, we calculated contrast from the observation image captured by the camera. And to evaluate characteristics of the eye-safe laser range scanner module, we reviewed the range measurement error under the same visibility constraint situations. In comparative tests, the CCD camera used is a general purpose model with a high magnification zoom lens. We evaluated a model with an effective measuring distance of 30 m, referring to the eye-safe laser range scanner module mounted on the Quince robot III. A visibility was introduced as quantitative evaluation parameter for comparing the monitoring performance of the color CCD camera and range measurement error of an eye-safe LRS module under aerosol environments where the fog particles are sprayed in the fog room. The visibility was calculated by attenuation of the He–Ne laser intensity in the fog room where the fog particles are being sprayed.
PubDate: 2019-03-26

• Handwritten Tifinagh Characters Recognition Using Deep Convolutional
Neural Networks
• Abstract: Handwritten characters recognition is a challenging task due to the diversity of writing styles. In the present paper, a new offline recognition system based on deep convolutional neural networks (CNNs) is designed. It uses CNNs to extract features from raw pixels which makes the proposed technique more flexible than the other conventional ones that require a set of additional preprocessing steps to extract the desired invariants features. The proposed system is tested on the AMHCD data set and achieved the best recognition accuracy (99.10%) compared to the state-of-the-art results.
PubDate: 2019-03-22

• A Mathematical Analysis Method of the Relationship Between DFT Magnitude
and Periodic Feature of a Signal
• Abstract: In this study, we developed a logical and complete mathematical analysis of the relationship between the discrete Fourier transform (DFT) magnitude and periodic feature of a signal. The physical meaning of the DFT magnitude index corresponds to the periodic signal feature; further analysis makes clear the relationship among DFT magnitude, magnitude index, and number of samples. The proposed analysis method also elucidates the relationship between alternating current magnitude index and frequency, by the way, the unit of image frequency will be given. The proposed analysis method is suitable for the machine vision and image processing in general.
PubDate: 2019-03-19

• An Effective Approach for Sub-acute Ischemic Stroke Lesion Segmentation by
Adopting Meta-Heuristics Feature Selection Technique Along with Hybrid
Naive Bayes and Sample-Weighted Random Forest Classification
• Abstract: In Earth, the most crucial illness for cause of death is ischemic stroke. Ischemic stroke arises as a result of an obstacle within a blood vessel supplying blood to the brain. In this paper, for sub-acute ischemic stroke lesion segmentation, we utilize an effective meta-heuristic feature selection technique along with hybrid Naive Bayes (NB) and sample weighted random forest (SWRF) classification approach. Initially, the features are extracted from the pre-processed image, after that, the feature selection is done by using the multi-objective enhanced firefly algorithm. To improve the classification performance, the dimensionality of the feature vectors and errors are reduced by eliminating such irrelevant and redundant features. After the feature selection process, an ensemble of NB and SWRF classifiers is used for segmenting the image. Here the NB classifier is trained and applied to estimate the weights of training samples. Then, the training samples with estimated weights are utilized to train SWRF. In our work stroke lesion segmentation is formulated as a binary classification problem where every local region is classified as either affected or non-affected area.
PubDate: 2019-03-08

• Decluttering Using Wavelet Based Higher Order Statistics and Target
Detection of GPR Images
• Authors: N. Smitha; Vipula Singh
Abstract: Images from ground penetrating radar (GPR) may be obscured by high clutter noise over the target signal, making target detection difficult. In this contribution, a decluttering technique is applied to GPR images. This clutter removal/reduction is achieved through wavelet decomposition and application of two methods using the third and fourth order statistics: skewness and kurtosis. These higher order statistics remove clutter but retain target signatures. Different scenarios are considered for real GPR images collected in our controlled lab environment set up and peak signal to noise ratio are compared for the two methods. Further features of targets and non-targets are extracted from de-noised images. These features are used in training a neural network classifier. This classifier is applied to various real GPR images with promising results for detection of targets.
PubDate: 2018-12-10
DOI: 10.1007/s11220-018-0223-x
Issue No: Vol. 20, No. 1 (2018)

• Modeling and Enhancement of Piezoelectric Accelerometer Relative
Sensitivity
• Authors: Salima Khaoula Reguieg; Zine Ghemari; Tarak Benslimane; Salah Saad
Abstract: The piezoelectric accelerometer is an electronic instrument based on the direct effect of the piezoelectric material, this device is widely used in the industries to monitor and detect defects of rotating machines in an early stage. In this paper, a thorough study of the piezoelectric accelerometer is carried out to understand its design and operation principle. A mathematical model of the accelerometer is developed based on Newton motion law then a new relative sensitivity equation in function of measurement error is extracted. This new equation has allowed a significant reduction in the measurement error, a maximum improvement in the precision and an optimization of the piezoelectric accelerometer relative sensitivity by the appropriate choice of damping rate. These improvements have optimized the accelerometer parameters and performances.
PubDate: 2018-12-04
DOI: 10.1007/s11220-018-0222-y
Issue No: Vol. 20, No. 1 (2018)

• On Stability Analysis of a Class of Nonlinear Systems with a Focus on
Composite Nonlinear Feedback Approach
• Authors: H. Ebrahimi Mollabashi; A. H. Mazinan
Abstract: Obtaining a desired transient performance of closed-loop system is the most important issues in impractical and industrial applications, in general. It is to note that the small overshoot and acceptable settling time of the system response are almost two typical characteristics in the transient performance. There are a number of contradictions between these performances; and therefore, a tradeoff between these ones should deeply be considered. In a word, a so-called composite nonlinear feedback (CNF) method is an efficient and simple technique which is employed to overcome the contradiction of simultaneous achievement of the mentioned transient performances. CNF based sliding mode control (SMC) for nonlinear systems with known upper bound uncertainty are considered in this paper. Unlike the existing results for the linear systems, the proposed CNF based SMC comprises two nonlinear parts. The first term assures the stability of the closed-loop nonlinear system and provides a fast convergence response. The second term reduces the overshoot of the response. Finally, to show the merits of the proposed approach, it is applied to two nonlinear Genesio’s chaotic system and a nonlinear helicopter system. The investigated results verify the effectiveness of the proposed approach.
PubDate: 2018-10-10
DOI: 10.1007/s11220-018-0221-z
Issue No: Vol. 19, No. 1 (2018)

• Plasmonic Biosensor in NIR with Chalcogenide Glass Material: On the Role
of Probe Geometry, Wavelength, and 2D Material
• Authors: Baljinder Kaur; Anuj K. Sharma
Abstract: Samarium doped chalcogenide core, perfluorinated polymer clad with Ag metal and plasmonic 2D materials based plasmonic fiber-optic sensor is simulated and analyzed in near-infrared (NIR) wavelength regime. Proposed sensor is directed at the detection of the malignancy stages of liver tissues. The performance analysis (in terms of figure-of-merit, i.e., FOM) was carried out taking into account the MoS2 and graphene monolayers as performance enhancing 2D materials. The analysis suggests that FOM values of MoS2-based sensor probe are better than graphene-based probe. Further, a comparative study shows that fiber-optic probe is able to provide much better performance than prism-based probes. The FOM gets better for longer wavelength. The specificity of the biosensor can be improved by employing a suitable buffer layer (1–15 nm) as a bio-recognition element.
PubDate: 2018-10-03
DOI: 10.1007/s11220-018-0220-0
Issue No: Vol. 19, No. 1 (2018)

• Color Image Modification with and without Hue Preservation
• Authors: Shashi Poddar; Marius Pedersen; Vinod Karar
Abstract: Color image modification is an essential component for several applications and the grayscale transformation is generally mapped to the color image indirectly. Although several techniques have been used for this transfer, they suffer from gamut mapping issue. In this paper, it is aimed to study the mapping of grayscale transformations to the color scale in different perspectives. Modifying the image in different color space than the original retains hue to a promising extent, but suffers from the gamut problem. A generic scheme to map grayscale changes to the color space for all kinds of spatial modification is proposed here. The hue preserving color image enhancement (HPCE) scheme discussed here is free of gamut-mapping issue and shows promising results in transferring the grayscale transformation to the color image in a simplistic manner. The proposed HPCE scheme is analysed qualitatively through visual appearance and quantitatively using color difference metrics SHAME and CID, gray image difference and EBCM measures. Different gray scale transformations such as S-type enhancement and different forms of histogram equalization techniques are applied on Berkeley dataset of 500 images to prove the efficacy of proposed algorithm.
PubDate: 2018-08-17
DOI: 10.1007/s11220-018-0219-6
Issue No: Vol. 19, No. 1 (2018)

• Recognizing Anomalies in Urban Road Scenes Through Analysing Single Images
Captured by Cameras on Vehicles
• Authors: Shuang Bai; Chao Han; Shan An
Abstract: In this paper, we propose to recognize anomalies in urban road scenes through analysing single images captured by cameras on driving vehicles. Anomaly detection is one of the most important functions for visual driver-assistance systems and autonomous vehicles. Anomaly detection provides drivers and autonomous vehicles important information about driving environments and help them drive more safely. In this work, we define anything on roads that are within a certain distance of a driving vehicle and pose potential dangers for it as anomalies, such as traffic accidents, reckless driven vehicles and pedestrians. The proposed approach recognizes anomalies in urban road scenes by analysing appearances of single images captured by cameras on driving vehicles. To do so, first, we collect a large number of urban road scene images that do not contain any anomalies. Second, we segment the road regions from these images and represent the obtained road regions based on the bag of visual words method. After that, we apply k-means clustering to the region representations for acquiring a small set of reference images. Third, we establish dense correspondence between input images and the reference images to create representations for the input images. Following that, representations of normal images are used to train a one-class Support Vector Machine classifier. Finally, we use the classifier to recognize images containing anomalies. Experiments on urban road scene images are conducted. Obtained results demonstrate that by using the proposed approach we can recognize urban road scene images containing anomalies.
PubDate: 2018-08-09
DOI: 10.1007/s11220-018-0218-7
Issue No: Vol. 19, No. 1 (2018)

• Lock-in Amplifier Based Eddy Current Instrument for Detection of
Sub-surface Defect in Stainless Steel Plates
• Authors: Anil Kumar Soni; B. Purnachandra Rao
Abstract: For detection of sub-surface defects using eddy current (EC) method, increasing the depth of penetration of ECs is essential. This can be achieved through strengthening of the primary magnetic field from the EC probe. This can be accomplished by using low-frequency high amplitude excitation current, precise phase lag measurement and high throughput probes. Working on these lines, the paper presents development of lock-in amplifier based EC instrument and cup-core send-receive type probe. Experiment results indicate that the proposed instrument and probe is able to detect sub-surface defects located at 8.0 mm below surface and classify sub-surface as well as surface defects in stainless steel plate.
PubDate: 2018-08-06
DOI: 10.1007/s11220-018-0217-8
Issue No: Vol. 19, No. 1 (2018)

• A Just Noticeable Difference-Based Video Quality Assessment Method with
Low Computational Complexity
• Authors: Woei-Tan Loh; David Boon Liang Bong
Abstract: A Just Noticeable Difference (JND)-based video quality assessment (VQA) method is proposed. This method, termed as JVQ, applies JND concept to structural similarity (SSIM) index to measure the spatial quality. JVQ incorporates three features, i.e. luminance adaptation, contrast masking, and texture masking. In JVQ, the concept of JND is refined and more features are considered. For the spatial part, minor distortions in the distorted frames are ignored and considered imperceptible. For the temporal part, SSIM index is simplified and used to measure the temporal video quality. Then, a similar JND concept which comprises of temporal masking is also applied in the temporal quality evaluation. Pixels with large variation over time are considered as not distorted because the distortions in these pixels are hardly perceivable. The final JVQ index is the arithmetic mean of both spatial and temporal quality indices. JVQ is found to achieve good correlation with subjective scores. In addition, this method has low computational cost as compared to existing state-of-the-art metrics.
PubDate: 2018-08-06
DOI: 10.1007/s11220-018-0216-9
Issue No: Vol. 19, No. 1 (2018)

JournalTOCs
School of Mathematical and Computer Sciences
Heriot-Watt University
Edinburgh, EH14 4AS, UK
Email: journaltocs@hw.ac.uk
Tel: +00 44 (0)131 4513762
Fax: +00 44 (0)131 4513327

Home (Search)
Subjects A-Z
Publishers A-Z
Customise
APIs