Subjects -> COMMUNICATIONS (Total: 518 journals)     - COMMUNICATIONS (446 journals)    - DIGITAL AND WIRELESS COMMUNICATION (31 journals)    - HUMAN COMMUNICATION (19 journals)    - MEETINGS AND CONGRESSES (7 journals)    - RADIO, TELEVISION AND CABLE (15 journals) DIGITAL AND WIRELESS COMMUNICATION (31 journals)
 Showing 1 - 31 of 31 Journals sorted alphabetically Ada : A Journal of Gender, New Media, and Technology       (Followers: 22) Advances in Image and Video Processing       (Followers: 24) Communications and Network       (Followers: 13) E-Health Telecommunication Systems and Networks       (Followers: 3) EURASIP Journal on Wireless Communications and Networking       (Followers: 14) Future Internet       (Followers: 84) Granular Computing IEEE Transactions on Wireless Communications       (Followers: 25) IEEE Wireless Communications Letters       (Followers: 41) IET Wireless Sensor Systems       (Followers: 17) International Journal of Communications, Network and System Sciences       (Followers: 9) International Journal of Digital Earth       (Followers: 14) International Journal of Embedded and Real-Time Communication Systems       (Followers: 9) International Journal of Interactive Communication Systems and Technologies       (Followers: 2) International Journal of Machine Intelligence and Sensory Signal Processing       (Followers: 3) International Journal of Mobile Computing and Multimedia Communications       (Followers: 2) International Journal of Satellite Communications and Networking       (Followers: 40) International Journal of Wireless and Mobile Computing       (Followers: 8) International Journal of Wireless Networks and Broadband Technologies       (Followers: 2) International Journals Digital Communication and Analog Signals       (Followers: 2) Journal of Digital Information       (Followers: 163) Journal of Interconnection Networks       (Followers: 1) Journal of the Southern Association for Information Systems       (Followers: 2) Mobile Media & Communication       (Followers: 10) Nano Communication Networks       (Followers: 5) Psychology of Popular Media Culture       (Followers: 2) Signal, Image and Video Processing       (Followers: 13) Ukrainian Information Space Vehicular Communications       (Followers: 4) Vista       (Followers: 2) Wireless Personal Communications       (Followers: 6)
Similar Journals
 Signal, Image and Video ProcessingJournal Prestige (SJR): 0.485 Citation Impact (citeScore): 2Number of Followers: 13      Hybrid journal (It can contain Open Access articles) ISSN (Print) 1863-1711 - ISSN (Online) 1863-1703 Published by Springer-Verlag  [2467 journals]
• A hybrid method for improving the luminosity and contrast of color retinal
images using the JND model and multiple layers of CLAHE

Abstract: Abstract Retinal imaging can be used to identify a variety of common eye and cardiac disorders. However, owing to non-uniform or poor illumination and low contrast, low-quality retinal fundus medical images are ineffective for diagnostic, particularly in computerized image analysis systems. The article proposes an effective image enhancement method for improving the luminosity and contrast of color retinal fundus images. To begin, the input color retinal fundus image is transformed to HSV (Hue, Saturation, and Value) color model, which separates the luminance channel (V) from the other color elements hue (H) and saturation (S). Then, on the luminosity channel (V), a new JND-based adaptive gamma correction method is utilized to improve the luminance of fundus images. After that, contrast is improved in the luminance component in the L*a*b* color space, employing a novel contrast enhancement technique that employs several layers of CLAHE (contrast limited adaptive histogram equalization). These two techniques substantially improve the overall luminance and contrast in retinal images while preserving the average brightness, keeping an original appearance, and maximizing the entropy of the input retinal fundus images. Experiments on a broad range of fundus images are performed to assess the proposed scheme's performance both qualitatively and quantitatively. Substantial objective evaluation indicates that the proposed scheme surpasses state-of-the-art enhancement techniques in terms of edge preservation index, entropy, a measure of enhancement, contrast ratio, and enhancement metrics. This retinal fundus image enhancement method can be employed to support ophthalmologists in effectively inspecting for retinal disorders and developing more accurate computerized image analysis for medical diagnosis.
PubDate: 2023-02-01

• Video quality enhancement using recursive deep residual learning network

Abstract: Abstract Outdoor images and videos suffer from several problems, such as the hazing problem due to the particles of dust, smoke, and other particles in the atmosphere. Videos in such atmospheric conditions are subject to visible quality degradations, such as low contrast and information loss. This paper presents a dehazing algorithm that enhances the video contrast, removes haze from hazy frames, and reduces frame degradation. We use a recursive deep residual learning (DRL) network as a dehazing tool to enhance video quality. The DRL network estimates the nonlinear mapping from the space of hazy input frames to that of output dehazed frames without estimating the transmission map and the atmospheric light as in traditional dehazing methods. After that, the dehazed frame is fed back to the input of the DRL network. This process is counted as an iteration. Our proposed algorithm depends on pre-processing of frames before the dehazing process to remove noise or enhance the visual quality, because all frames contain some noise due to sensor measurement errors. Noise can be amplified in the haze removal process if ignored. We use different types of enhancement techniques before the dehazing process. In addition, we modify the DRL network to be suitable for both near infrared (NIR) and visible frames. The number of iterations in the DRL network is increased from three iterations to nine iterations and the effect of increasing the number of iterations on the output dehazed frames is studied. We stopped at nine iterations, because elapsed time increases with the increase in the number of iterations. The peak signal-to-noise ratio and correlation after the dehazing process between dehazed and input hazy frames are used as evaluation metrics for our proposed algorithm. Results show that by increasing the number of iterations in the DRL network, dehazed frames record the best contrast, the highest spectral entropy, and the highest visual quality.
PubDate: 2023-02-01

• Study on steel plate scratch detection based on improved MSR and phase
consistency

Abstract: Abstract In the casting process of the steel plate, due to the influence of rolling equipment and technology, the defects such as cracks and scratches appear on the surface of steel plate, which affect the performance of steel plate and even cause production accidents. In this paper, an automatic detection method for steel plate scratch is proposed. Firstly, the steel plate image is decomposed by channel and the enhanced image is obtained by the improved MSR (Multi-Scale Retinex) enhancement algorithm. Then, the phase consistency is detected after the Log Gabor wavelet transform and the scratch areas are obtained by the threshold segmentation and intersection of three channels. Finally, the scratch position is identified and the scratch characteristics such as width and length can be calculated. The results show that the minimum error of the characteristics measurement is only 2.28% in the experimental environment and 4.15% in the field environment, and the mean running time is 0.2826 s in the experimental environment and 0.3193 s in the field environment. It verifies that the proposed method is effective and practical.
PubDate: 2023-02-01

• Two-stage deep learning framework for sRGB image white balance

Abstract: Abstract This work aims to correct white-balance errors in sRGB images. These white-balance errors are hard to fix due to the nonlinear color-processing procedures applied by camera image signal processors (ISP) to produce the final sRGB colors. Camera ISPs apply these nonlinear procedures after the essential white-balance step to render sensor raw images to the sRGB space through a camera-specific set of tone curves and look-up tables. To correct improperly white-balanced images, projecting non-linear sRGB colors back to their original raw space is required. Recent work formulates the problem as an image translation problem, where input sRGB colors are mapped using nonlinear polynomial correction functions to fix such white-balance errors. In this work, we show that correcting white-balance errors in sRGB images through a global color mapping followed by spatially local adjustments, learned in an end-to-end training, introduces perceptual improvements in the final results. Qualitative and quantitative comparisons with recently published methods for camera-rendered image white balancing validate our method’s efficacy and show that our method achieves competitive results with state-of-the-art methods.
PubDate: 2023-02-01

• A novel mixing matrix estimation method for underdetermined blind source
separation based on sparse subspace clustering

Abstract: Abstract It is essential to accurately estimate the mixing matrix and determine the number of source signals in the problem of underdetermined blind source separation. The problem is solved in this paper via sparse subspace clustering, which can be used to found low-dimensional data structures in observed data. To enhance the linear clustering characteristics of time-frequency points, the high energy points are reserved first, and the angle difference of real and imaginary portions is employed to screen single source points. After that, the time-frequency points are clustered using sparse subspace clustering, and the number of source signals is identified. Finally, the local density of eigenvectors is used to determine the mixing matrix. The proposed algorithm is capable of accurately estimating the mixing matrix. It has strong robustness and adaptable to a wide range of mixing circumstances. The proposed method’s effectiveness is demonstrated by theoretical analysis and experimental data.
PubDate: 2023-02-01

• Comparative analysis of post-processing on spectral collocation methods
for non-smooth functions

Abstract: Abstract Discretization-based spectral approximation methods provide spectrally accurate reconstruction of an analytic function. The expansion of non-smooth functions is contaminated by high frequency non-diminishing oscillations near discontinuity points, and this behaviour is named as Gibbs phenomenon. This problem can be well resolved by well-chosen post-processing technique, and one possible choice is spectral filtering. In this paper, a comparison scenario of adaptive spectral filtering for resolution of Gibbs phenomenon is presented. Several spectral filter functions are compared using Chebyshev collocation and Legendre collocation spectral methods, in terms of pointwise and $$L_2$$ normed-wise convergence analysis of computed filtered approximations.
PubDate: 2023-02-01

• Crack image recognition on fracture mechanics cross valley edge detection
by fractional differential with multi-scale analysis

Abstract: Abstract The recognition of pavement cracks is crucial in road engineering and airport maintains. In order to successfully apply image processing technique for automatic crack detection, the first and hardest task is to recognize crack images in a huge number of pavement images. To do this, the image processing technique and Fracture mechanics are combined first time in this area, the studied method includes four steps: (1) The pavement crack image shrinking is carried out by a proposed multi-scale analysis algorithm, which is more effective for both preserving weak valley edges and reducing computing cost; (2) Then, a so called valley edge detection algorithm based on Fractional differential for finding local dark line/curve is studied for tracing crack segments, it considers template size, weighted average gray level value in each line in four different directions, the output can be a gradient magnitude image or a binary image; (3) In the binary image, the crack segments are refined based on a number of post processing functions to remove noise and fill segment gaps; and (4) After that, to quickly judge if the image has cracks, Fracture mechanics is applied to calculate the judgment parameter T, which is directly proportion to the image edge density, and the ratio between the average gradient magnitude value and the average gray level value in the candidate crack segment. In experiments, more than 400 pavement images (the resolution is 4096 × 2048 pixels) are tested, and the crack identification accuracy is up to 97%.
PubDate: 2023-02-01

• A least-squares method for simultaneous synchronization and relative
calibration of overlapped videos

Abstract: Abstract In this paper, a straightforward mathematical model is proposed to synchronize and estimate the relative parameters of videos taken with a fixed relative orientation. The foundation of this model was the well-known coplanarity condition that prevails between matched points of two perspective images. Nevertheless, the synchronization problem has also been incorporated into it by making the matched points dependent on time. In this method, the required control data provides by tracking the positions of moving points in the temporal and spatial overlaps of the videos. Also, the unknown parameters are estimated through the least-squares estimation of a constrained system of linearized equations. The results of implementations on different datasets have demonstrated the efficiency of the proposed method in the temporal and relative calibration of stereo videos; as it has reached on average to the one frame accuracy in synchronization and 4.3 pixels precision in generalization of relative calibrations.
PubDate: 2023-02-01

• Fluorescence microscopy image noise reduction using IEMD-based adaptive
thresholding approach

Abstract: Abstract Fluorescence microscopy is an important investigation tool for discoveries in the field of biological sciences. In this paper, we propose an adaptive thresholding technique-based improved empirical mode decomposition (IEMD) for denoising of heavily degraded images labeled with Fluorescent proteins. These images are widely used by a computational biologists to analyze the biological functions of different species. A variance stabilization transformation is applied as preprocessing step. The multi-scale Wiener filtering approach is used as the first step for accurate image deconvolution. In the subsequent steps, IEMD is performed to obtain different series of intrinsic mode functions (IMFs) which are further separated into noise and signal-significant IMFs based on Cosine similarity index. The IMF adaptive thresholding technique is used which filter-out the unwanted frequency coefficients related to mixed Poisson–Gaussian noise (MPG). The thresholded output IMFs are combined with signal significant IMFs in the third step. Finally, the mean square deviation (MSD) is minimized using mixed Poisson–Gaussian unbiased risk estimate (MPGURE). To evaluate the effectiveness of the proposed scheme, we have compared the results of the proposed scheme with those of the five state-of-the-art techniques. The simulation results validate, the effectiveness of the proposed method. The proposed algorithm achieves better performance in terms of four quantitative evaluation measures by reducing the effect of noise.
PubDate: 2023-02-01

• One-dimensional block-matching motion estimation algorithm

Abstract: Abstract Motion estimation is a fundamental problem in the field of video restoration. The traditional two-dimensional block-matching algorithm has better search quality, but the search speed is slow. Based on the two-dimensional block-matching algorithm, we perform dimensionality reduction processing on the matching block and propose a one-dimensional block-matching motion estimation algorithm. According to the error of the motion estimation of the video sequence to be repaired, we establish a new motion estimation model to discuss the reasons for the deviation of the motion estimation. The experimental results show that compared with the traditional two-dimensional block-matching motion estimation algorithm, the one-dimensional block-matching motion estimation algorithm can significantly calculate the search speed under the premise of ensuring the accuracy of the estimation. In the experiments with multiple impairments added, the stability of the one-dimensional block-matching motion estimation algorithm is better than that of the two-dimensional block-matching motion estimation algorithm.
PubDate: 2023-02-01

• A one-stage deep learning framework for automatic detection of safety
harnesses in high-altitude operations

Abstract: Abstract Safety harness plays an essential role in protecting the workers in high-altitude operations from falls from heights. Automatic detection of safety harness wearing is significant for safety management. To deal with the inherent problems of the existing two-stage detection method for safety harnesses, a novel one-stage detection framework is designed by incorporating several promising modules into a YOLO network, which is end-to-end trained. Here, the dilated convolution module and the depth-wise separable convolution module are subsequently incorporated to improve the overall receptive fields of feature maps and to reduce the amount of calculation, respectively. An attention proposal sub-network (APN) is introduced for fine-grained feature learning. To improve the convergence of the proposed framework, a novel loss function is designed by adding a penalty term into the loss function named complete intersection over union (CIoU). Also, to facilitate the study, a new and publicly available dataset for safety harness wearing detection is constructed, which consists of 2617 images including 8163 safety harness examples. Experimental results demonstrate that the proposed framework can perform an excellent task for safety harness wearing detection with 80.25% mAP at a reasonable speed of 29.18 FPS, especially for small instances.
PubDate: 2023-02-01

• An improved polynomial rooting-based method for solving non-trivial
ambiguity in direction-finding using an unfolded co-prime linear array

Abstract: Abstract The direction-finding (DF) problem for unfolded co-prime linear array (UCLA) is researched. Specifically, there is a need to address the critical issue of non-trivial ambiguity in estimating the angle-of-arrival (AOA) parameter. To address this issue, an improved polynomial rooting-based method is proposed. A polynomial function is derived based on the orthogonality between the noise subspace singular vectors and array response vectors. In order to select the signal roots that are related to true AOAs over ambiguous roots, a maximum signal power function is proposed based on spatial filtering and second-order differential. The proposed method overcomes the non-trivial ambiguity and estimates the true AOAs successfully with improved estimation performances in terms of reliability, accuracy and angular resolution involving low computational cost. Simulations have been performed to show the effectiveness and superiority of the proposed method.
PubDate: 2023-02-01

• Local channel transformation for efficient convolutional neural network

Abstract: Abstract The efficiency of the convolutional neural network (CNN) model is one of the biggest limitations when CNN is applied to mobile devices. As an efficient convolution method, the pointwise convolution is widely used in many networks to achieve the expansion and compression of channel dimensions. Nevertheless, pointwise convolution still needs to consume a vast number of calculations and parameters. In this paper, depthwise channel ascent (DCA) and group channel descent (GCD) were introduced as efficient channel transformation methods to replace pointwise convolution. DCA and GCD decompose the global channel transformation into local channel transformation. DCA utilizes spatial features to expand channel dimension, while GCD utilizes non-learning channel compression to decrease channel dimension. Compared with other counterparts, networks equipped with DCA and GCD can significantly reduce the calculations and parameters while maintaining competitive accuracy. The performance of the proposed method had been verified in different networks and multiple datasets. In the DCASE2019 dataset, the proposed method can reduce the parameters and calculation of MobileNetV2 to 41.9% and 30.2% and accelerate the inference speed to 73.8%.
PubDate: 2023-02-01

• Semantic-oriented learning-based image compression by Only-Train-Once
quantized autoencoders

Abstract: Abstract Accessibility to big training datasets together with current advances in computing power has emerged interest in the leverage of deep learning to address image compression. This needs to train and deploy separate networks for rate adaptation, which is impractical and extensive in terms of memory cost and power consumption, especially for broad bitrate ranges. To deal with such limitation, the variable-rate compression methods use the Lagrange multiplier to control the Rate/Distortion trade-offs in order not to require retraining of the neural network for each rate. However, they do not make an optimized bit allocation for the eye-catching foreground details, and do not consider the different degree of attention that the human eye has to each area of the image. Thus, other deep learning-based image compression approaches, which could outperform the above ones, are replied on the use of additional information. In this paper, we present a loss-conditional autoencoder tailored to the specific task of semantic image understanding to achieve higher visual quality in lossy variable-rate compression. Our framework is a neural network-based scheme able to automatically optimize coding parameters with multi-term perceptual loss function based on semantic-important structural SIMilarity index. To ensure the rate adaptation, we suggest modulating the compression network on the bitwidth of its activations by quantizing them according to several bitwidth values. Experiments are presented on the JPEG AI dataset in which our method achieves competitive and higher visual quality for the same compressed size, when compared to conventional codecs and related work.
PubDate: 2023-02-01

• Focus on local: transmission line defect detection via feature refinement

Abstract: Abstract Different from the object detection which has made great progress in natural imagery, transmission line images acquired by UAVs have their own challenges in detecting defects in critical parts, such as object scale variation and small defect targets. In this paper, we construct an effective architecture, called FOLO, to improve the accuracy of defect detection in critical parts of transmission lines. To capture the critical part defect object features, a local contextual feature pyramid network (LCFPN) is proposed to refine the local contextual information and perform multi-scale learning. In LCFPN, we introduce a channel feature refinement block (CFRB) and multiple spatial feature refinement block (SFRBs) to further improve the ability of the network to focus on local features. Besides, a local adaptive feature network (LAFN) is designed, which makes it possible to locate adaptive components with defects in critical areas of different shapes. Since existing transmission line datasets have a single category, we create a new defect detection dataset containing insulators, anti-vibration hammers and bird nests, named IVB. Experimental results on IVB show that the proposed FOLO yields promising performance against other approaches.
PubDate: 2023-02-01

• Architectural style classification based on CNN and channel–spatial
attention

Abstract: Abstract The accurate classification of architectural styles is of great significance to the study of architectural culture and human historical civilization. Models based on convolutional neural network (CNN) have achieved highly competitive results in the field of architectural style classification owing to its more powerful capability of feature expression. However, most of the CNN models to date only extract the global features of architecture facade or focus on some regions of architecture and fail to extract the spatial features of different components. To improve the accuracy of architectural style classification, we propose an architectural style classification method based on CNN and channel–spatial attention. Firstly, we add a preprocessing operation before CNN feature extraction to select main building candidate region in architectural image and then use CNN feature extractor for deep feature extraction. Secondly, channel–spatial attention module is introduced to generate an attention map, which can not only enhance the texture feature representation of architectural images but also focus on the spatial features of different architectural elements. Finally, the Softmax classifier is used to predict the score of the target class. The experimental results on the Architectural Style Dataset and AHE_Dataset have achieved satisfactory performance.
PubDate: 2023-02-01

• An efficient approach for image de-fencing based on conditional generative

Abstract: Abstract Automated image de-fencing is an important area of computer vision that deals with the problem of virtually removing fence structures, if any, from images and produce aesthetically pleasing images without the fence structures. Unlike most of the previous de-fencing approaches that employ a two-stage process of fence mask detection followed by image inpainting, here we present a single-stage end-to-end conditional generative adversarial network-based de-fencing model that takes as input a fenced image and produces the corresponding de-fenced image in only 16 ms. The proposed network has been trained using an extensive dataset of fenced and ground-truth de-fenced image pairs by employing a combination of adversarial loss, L1 loss, perceptual loss, and estimated fence mask loss till convergence. The experimental results shows that our approach is capable of successfully handling images with even broken, irregular, and occluded fence structures. Qualitative and quantitative comparative study with previous de-fencing methods also show that our approach outperforms these existing techniques in terms of both response time and quality of de-fencing.
PubDate: 2023-02-01

• Diagnosis of schizophrenia using brain resting-state fMRI with activity
maps based on deep learning

Abstract: Abstract In recent years, the use of medical imaging in the analysis of human body structure and diagnosis of diseases has greatly increased. Schizophrenia is a serious psychiatric illness which needs early and accurate diagnosis. Functional magnetic resonance imaging (fMRI) with appropriate spatial resolution is a powerful technique for visualizing human brain activity. One of the major challenges in classifying these images is the high-dimensional fMRI images along with the poor quality of these data. In this study, a general framework for classifying images into two groups of healthy and schizophrenic patients is presented. In this method, after preprocessing fMRI images, functional connectivity analysis is used to extract the features. After extracting functional mappings, we use three-dimensional convolutional neural network and long short-term memory recurrent network to extract spatial and temporal information to classify activity maps. Our results show that the use of these features can lead to a strong classification on the COBRE dataset with an accuracy of 92.32%, which is very promising.
PubDate: 2023-02-01

• Action recognition based on attention mechanism and depthwise separable
residual module

Abstract: Abstract Aiming at the deficiencies of the lightweight action recognition network YOWO, a dual attention mechanism is proposed to improve the performance of the network. It is further proposed to use the depthwise separable convolution to replace part of the ordinary convolution after the 2D and 3D fusion and use residuals to merge the feature maps of different convolutional layers of the network, which improves the performance and speed of the network. First, to more effectively obtain salient features from the space and channel dimensions, add the CBAM space and channel attention module to the network; then, to make the parameters of the network more lightweight, it is proposed to use depthwise separable convolution to replace part of the ordinary convolution in the YOWO network. From experiments on the UCF101-24 and J-HMDB-21 datasets, compared with YOWO network, the improved method has significantly improved the accuracy and speed of action recognition.
PubDate: 2023-02-01

• End-to-end deep learning of lane detection and path prediction for
real-time autonomous driving

Abstract: Abstract Inspired by the UNet architecture of semantic image segmentation, we propose a lightweight UNet using depthwise separable convolutions (DSUNet) for end-to-end learning of lane detection and path prediction (PP) in autonomous driving. We also design and integrate a PP algorithm with convolutional neural network (CNN) to form a simulation model (CNN-PP) that can be used to assess CNN’s performance qualitatively, quantitatively, and dynamically in a host agent car driving along with other agents all in a real-time autonomous manner. DSUNet is 5.12 $$\times$$ lighter in model size and 1.61 $$\times$$ faster in inference than UNet. DSUNet-PP outperforms UNet-PP in mean average errors of predicted curvature and lateral offset for path planning in dynamic simulation. DSUNet-PP outperforms a modified UNet in lateral error, which is tested in a real car on real road. These results show that DSUNet is efficient and effective for lane detection and path prediction in autonomous driving.
PubDate: 2023-02-01

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