Search

Due to problems such as the shooting light, viewing angle, and camera equipment, low-light images with low contrast, color distortion, high noise, and unclear details can be seen regularly in real scenes. These low-light images will not only affect our observation but will also greatly affect the performance of computer vision processing algorithms. Low-light image enhancement technology can help to improve the quality of images and make them more applicable to fields such as computer vision, machine learning, and artificial intelligence. In this paper, we propose a novel method to enhance images through Bézier curve estimation. We estimate the pixel-level Bézier curve by training a deep neural network (BCE-Net) to adjust the dynamic range of a given image. Based on the good properties of the Bézier curve, in that it is smooth, continuous, and differentiable everywhere, low-light image enhancement through Bézier curve mapping is effective. The advantages of BCE-Net’s brevity and zero-reference make it generalizable to other low-light conditions. Extensive experiments show that our method outperforms existing methods both qualitatively and quantitatively.

With the advent of the 5G era, radio frequency identification (RFID) has been widely applied in various fields as one of the key technologies for the Internet of Things (IoT) to realize the Internet of Everything (IoE). In recent years, RFID-based motion sensing has emerged as an important research area with great potential for development. In this paper, an RFID backscatter sport motion sensing scheme is proposed, which effectively solves the multi-classification problem by using the received signal strength (RSS) of the backscattered RFID and the error correcting output coding (ECOC)-based support vector machine (SVM). We conduct extensive experiments to validate the effectiveness of the proposed scheme, in which the signal intensities of different types of action poses are collected and the SVM is used as the classification algorithm to achieve high classification accuracies.

The differences in urinary proteins could provide a novel opportunity to distinguish the different types of drug-induced kidney injury (DIKI). In this research, Au nanoparticles–polyethyleneimine (AuNPs–PEI) and the three fluorophore-labeled proteins (FLPs) have been constructed as a multichannel fluorescent array sensor via electrostatic interaction, which was used to detect the subtle changes in urine collected from the pathological state of DIKI. Once the urine from different types of DIKI was introduced, the binding equilibrium between AuNPs–PEI and FLPs would be broken due to the competitive binding of urinary protein, and the corresponding fluorescence response pattern would be generated. Depending on the different fluorescence response patterns, the different types of DIKI were successfully identified by principal component analysis (PCA) and linear discriminant analysis (LDA). Accordingly, the strategy was expected to be a powerful technique for evaluating the potential unclear mechanisms of nephrotoxic drugs, which would provide a promising method for screening potential renal-protective drugs.

Remote sensing images often have limited resolution, which can hinder their effectiveness in various applications. Super-resolution techniques can enhance the resolution of remote sensing images, and arbitrary resolution super-resolution techniques provide additional flexibility in choosing appropriate image resolutions for different tasks. However, for subsequent processing, such as detection and classification, the resolution of the input image may vary greatly for different methods. In this paper, we propose a method for continuous remote sensing image super-resolution using feature-enhanced implicit neural representation (SR-FEINR). Continuous remote sensing image super-resolution means users can scale a low-resolution image into an image with arbitrary resolution. Our algorithm is composed of three main components: a low-resolution image feature extraction module, a positional encoding module, and a feature-enhanced multi-layer perceptron module. We are the first to apply implicit neural representation in a continuous remote sensing image super-resolution task. Through extensive experiments on two popular remote sensing image datasets, we have shown that our SR-FEINR outperforms the state-of-the-art algorithms in terms of accuracy. Our algorithm showed an average improvement of 0.05 dB over the existing method on ×30 across three datasets.

Keeping railway tracks in good operational condition is one of the most important tasks for railway owners. As a result, railway companies have to conduct track inspections periodically, which is costly and time-consuming. Due to the rapid development in computer science, many prediction models using machine learning methods have been developed. It is possible to discover the degradation pattern and develop accurate prediction models. The paper reviews the existing prediction methods for railway track degradation, including traditional methods and prediction methods based on machine learning methods, including probabilistic methods, Artificial Neural Network (ANN), Support Vector Machine (SVM), and Grey Model (GM). The advantages, shortage, and applicability of methods are discussed, and recommendations for further research are provided.

The fiber optic gyroscope (FOG) is a high precision inertial navigation device, and it is necessary to ensure its reliability for effective use. However, the extracted fault features are easily distorted due to the interference of vibrations when the FOG is in operation. In order to minimize the influence of vibrations to the greatest extent, a fusion diagnosis method was proposed in this paper. It extracted features from fault data with Fast Fourier Transform (FFT) and wavelet packet decomposition (WPD), and built a strong diagnostic classifier with a sparse auto encoder (SAE) and a neural network (NN). Then, a fusion neural network model was established based on the diagnostic output probabilities of the two primary classifiers, which improved the diagnostic accuracy and the anti-vibration capability. Then, five fault types of the FOG under random vibration conditions were established. Fault data sets were collected and generated for experimental comparison with other methods. The results showed that the proposed fusion fault diagnosis method could perform effective and robust fault diagnosis for the FOG under vibration conditions with a high diagnostic accuracy.

Depth estimation based on light field imaging is a new methodology that has succeeded the traditional binocular stereo matching and depth from monocular images. Significant progress has been made in light-field depth estimation. Nevertheless, the balance between computational time and the accuracy of depth estimation is still worth exploring. The geometry in light field imaging is the basis of depth estimation, and the abundant light-field data provides convenience for applying deep learning algorithms. The Epipolar Plane Image (EPI) generated from the light-field data has a line texture containing geometric information. The slope of the line is proportional to the depth of the corresponding object. Considering the light field depth estimation as a spatial density prediction task, we design a convolutional neural network (ESTNet) to estimate the accurate depth quickly. Inspired by the strong image feature extraction ability of convolutional neural networks, especially for texture images, we propose to generate EPI synthetic images from light field data as the input of ESTNet to improve the effect of feature extraction and depth estimation. The architecture of ESTNet is characterized by three input streams, encoding-decoding structure, and skipconnections. The three input streams receive horizontal EPI synthetic image (EPIh), vertical EPI synthetic image (EPIv), and central view image (CV), respectively. EPIh and EPIv contain rich texture and depth cues, while CV provides pixel position association information. ESTNet consists of two stages: encoding and decoding. The encoding stage includes several convolution modules, and correspondingly, the decoding stage embodies some transposed convolution modules. In addition to the forward propagation of the network ESTNet, some skip-connections are added between the convolution module and the corresponding transposed convolution module to fuse the shallow local and deep semantic features. ESTNet is trained on one part of a synthetic light-field dataset and then tested on another part of the synthetic light-field dataset and real light-field dataset. Ablation experiments show that our ESTNet structure is reasonable. Experiments on the synthetic light-field dataset and real light-field dataset show that our ESTNet can balance the accuracy of depth estimation and computational time.

In a tunneling boring machine (TBM), to obtain the attitude in real time is very important for a driver. However, the current laser targeting system has a large delay before obtaining the attitude. So, an adaptive-neuro-fuzzy-based information fusion method is proposed to predict the attitude of a laser targeting system in real time. In the proposed method, a dual-rate information fusion is used to fuse the information of a laser targeting system and a two-axis inclinometer, and then obtain roll and pitch angles with a higher rate and provide a smoother attitude prediction. Considering that a measurement error exists, the adaptive neuro-fuzzy inference system (ANFIS) is proposed to model the measurement error, and then the ANFIS-based model is combined with the dual-rate information fusion to achieve high performance. Experimental results show the ANFIS-based information fusion can provide higher real-time performance and accuracy of the attitude prediction. Experimental results also verify that the ANFIS-based information fusion can solve the problem of the laser targeting system losing signals.

In this paper, multi-physical models of bending characteristics, including the static, dynamic and microwave models, are firstly proposed for the double-clamped beam switch based on flexible substrate. Both simulated and experimental verification have been carried out to prove that the changing regularity of the driving voltage and time of the switch is inversely proportional with the increase in the bending curvature of the flexible substrate. The microwave performance of the switch at the ON state is found to get worse with the increase in the bending curvature. The measured results indicate that when the bending curvature increases from 0 m−1 to 28.6 m−1, the measured driving voltage decreases from 90.0 V to 72.6 V with the error of 5.9% compared with the calculated results. The measured driving time decreases from 52.4 μs to 35.6 μs with the error of 16.7% compared with the calculated results. When the substrate bending curvature increases from 0 m−1 to 28.6 m−1, the measured reflection loss S11 of the switch gradually deteriorates from −27.1 dB to −22.0 dB with the error of 1.3 dB corresponding to the calculated results at 10 GHz. All the simulated and experimental results are consistent with the theoretical calculated results.

Surface plasmon resonance (SPR) with two-dimensional (2D) materials is proposed to enhance the sensitivity of sensors. A novel Goos–Hänchen (GH) shift sensing scheme based on blue phosphorene (BlueP)/transition metal dichalogenides (TMDCs) and graphene structure is proposed. The significantly enhanced GH shift is obtained by optimizing the layers of BlueP/TMDCs and graphene. The maximum GH shift of the hybrid structure of Ag-Indium tin oxide (ITO)-BlueP/WS2–graphene is −2361λ with BlueP/WS2 four layers and a graphene monolayer. Furthermore, the GH shift can be positive or negative depending on the layer number of BlueP/TMDCs and graphene. For sensing performance, the highest sensitivity of 2.767 × 107λ/RIU is realized, which is 5152.7 times higher than the traditional Ag-SPR structure, 2470.5 times of Ag-ITO, 2159.2 times of Ag-ITO-BlueP/WS2, and 688.9 times of Ag-ITO–graphene. Therefore, such configuration with GH shift can be used in various chemical, biomedical and optical sensing fields.

In order to improve the performance of surface plasmon resonance (SPR) biosensor, the structure based on two-dimensional (2D) of graphene and transition metal dichalcogenides (TMDCs) are proposed to greatly enhance the Goos-Hänchen (GH) shift. It is theoretically proved that GH shift can be significantly enhanced in SPR structure coated with gold (Au)-indium tin oxide (ITO)-TMDCs-graphene heterostructure. In order to realize high GH shifts, the number of TMDCs and graphene layer are optimized. The highest GH shift (−801.7 λ) is obtained by Au-ITO-MoSe2-graphene hybrid structure with MoSe2 monolayer and graphene bilayer, respectively. By analyzing the GH variation, the index sensitivity of such configuration can reach as high as 8.02 × 105 λ/RIU, which is 293.24 times of the Au-ITO structure and 177.43 times of the Au-ITO-graphene structure. The proposed SPR biosensor can be widely used in the precision metrology and optical sensing.

This study investigated a novel method of fusing visible (VIS) and infrared (IR) images with the major objective of obtaining higher-resolution IR images. Most existing image fusion methods focus only on visual performance and many fail to consider the thermal physical properties of the IR images, leading to spectral distortion in the fused image. In this study, we use the IR thermal physical property to correct the VIS image directly. Specifically, the Stefan-Boltzmann Law is used as a strong constraint to modulate the VIS image, such that the fused result shows a similar level of regional thermal energy as the original IR image, while preserving the high-resolution structural features from the VIS image. This method is an improvement over our previous study, which required VIS-IR multi-wavelet fusion before the same correction method was applied. The results of experiments show that applying this correction to the VIS image directly without multi-resolution analysis (MRA) processing achieves similar results, but is considerably more computationally efficient, thereby providing a new perspective on VIS and IR image fusion.

Geostationary meteorological satellite infrared (IR) channel data contain important spectral information for meteorological research and applications, but their spatial resolution is relatively low. The objective of this study is to obtain higher-resolution IR images. One common method of increasing resolution fuses the IR data with high-resolution visible (VIS) channel data. However, most existing image fusion methods focus only on visual performance, and often fail to take into account the thermal physical properties of the IR images. As a result, spectral distortion occurs frequently. To tackle this problem, we propose a thermal physical properties-based correction method for fusing geostationary meteorological satellite IR and VIS images. In our two-step process, the high-resolution structural features of the VIS image are first extracted and incorporated into the IR image using regular multi-resolution fusion approach, such as the multiwavelet analysis. This step significantly increases the visual details in the IR image, but fake thermal information may be included. Next, the Stefan-Boltzmann Law is applied to correct the distortion, to retain or recover the thermal infrared nature of the fused image. The results of both the qualitative and quantitative evaluation demonstrate that the proposed physical correction method both improves the spatial resolution and preserves the infrared thermal properties.

We investigated tip-enhanced Raman spectra excited by high-order fiber vector beams. Theoretical analysis shows that the high-order fiber vector beams have stronger longitudinal electric field components than linearly polarized light under tight focusing conditions. By introducing the high-order fiber vector beams and the linearly polarized beam from a fiber vector beam generator based on an electrically-controlled acoustically-induced fiber grating into a top-illumination tip-enhanced Raman spectroscopy (TERS) setup, the tip-enhanced Raman signal produced by the high-order fiber vector beams was 1.6 times as strong as that produced by the linearly polarized light. This result suggests a new type of efficient excitation light beams for TERS.

Coalbed methane (CBM) is one kind of clean-burning gas and has been valued as a new form of energy that will be used widely in the near future. When producing CBM, the working level within a CBM wellbore annulus needs to be monitored to dynamically adjust the gas drainage and extraction processes. However, the existing method of measuring the working level does not meet the needs of accurate adjustment, so we designed a new sensor for this purpose. The principle of our sensor is a liquid pressure formula, i.e., the sensor monitors the two-phase flow patterns and obtains the mean density of the two-phase flow according to the pattern recognition result in the first step, and then combines the pressure data of the working level to calculate the working level using the liquid pressure formula. The sensor was tested in both the lab and on site, and the tests showed that the sensor’s error was ±8% and that the sensor could function well in practical conditions and remain stable in the long term.

The measurement of wellbore annulus gas–liquid two-phase flow in CBM (coalbed methane) wells is of great significance for reasonably developing gas drainage and extraction processes, estimating CBM output, judging the operating conditions of CBM wells and analyzing stratum conditions. Hence, a specially designed sensor is urgently needed for real-time measurement of gas–liquid two-phase flow in CBM wellbore annulus. Existing flow sensors fail to meet the requirements of the operating conditions of CBM wellbore annulus due to such factors as an inapplicable measurement principle, larger size, poor sealability, high installation accuracy, and higher requirements for fluid media. Therefore, based on the principle of a target flowmeter, this paper designs a new two-phase flow sensor that can identify and automatically calibrate different flow patterns of two-phase flows. Upon the successful development of the new flow sensor, lab and field tests were carried out, and the results show that the newly designed sensor, with a measurement accuracy of ±2.5%, can adapt to the operating conditions of CBM wells and is reliable for long-term work.

The parameters of gas-liquid two-phase flow bubbles in field coalbed methane (CBM) wells are of great significance for analyzing coalbed methane output, judging faults in CBM wells, and developing gas drainage and extraction processes, which stimulates an urgent need for detecting bubble parameters for CBM wells in the field. However, existing bubble detectors cannot meet the requirements of the working environments of CBM wells. Therefore, this paper reports findings on the principles of measuring the flow pattern, velocity, and volume of two-phase flow bubbles based on conductivity, from which a new bubble sensor was designed. The structural parameters and other parameters of the sensor were then computed, the “water film phenomenon” produced by the sensor was analyzed, and the appropriate materials for making the sensor were tested and selected. After the sensor was successfully devised, laboratory tests and field tests were performed, and the test results indicated that the sensor was highly reliable and could detect the flow patterns of two-phase flows, as well as the quantities, velocities, and volumes of bubbles. With a velocity measurement error of ±5% and a volume measurement error of ±7%, the sensor can meet the requirements of field use. Finally, the characteristics and deficiencies of the bubble sensor are summarized based on an analysis of the measurement errors and a comparison of existing bubble-measuring devices and the designed sensor.

With the gradual popularization of Internet-of-Things (IoT) applications and the development of wireless networking technologies, the use of heterogeneous devices and runtime verification of task fulfillment with different constraints are required in real-world IoT scenarios. As far as IoT systems are concerned, most of them are built on centralized architectures, which reveal various assailable points in data security and privacy threats. Hence, this paper aims to investigate these issues by delegating the responsibility of a verification monitor from a centralized architecture to a decentralized manner using blockchain technology. We present a smart contract-based task management scheme to provide runtime verification of device behaviors and allows trustworthy access control to these devices. The business logic of the proposed system is specified by the smart contract, which automates all time-consuming processes cryptographically and correctly. The usability of the proposed solution is further demonstrated by implementing a prototype application in which the Hyperledger Fabric is utilized to implement the business logic for runtime verification and access control with one desktop and one Raspberry Pi. A comprehensive evaluation experiment is conducted, and the results indicate the effectiveness and efficiency of the proposed system.

With the rapid development of communication technologies, the Internet of Things (IoT) is getting out of its infancy, into full maturity, and tends to be developed in an explosively rapid way, with more and more data transmitted and processed. As a result, the ability to manage devices deployed worldwide has been given more and advanced requirements in practical application performances. Most existing IoT platforms are highly centralized architectures, which suffer from various technical limitations, such as a cyber-attack and single point of failure. A new solution direction is essential to enhance data accessing, while regulating it with government mandates in privacy and security. In this paper, we propose an integrated IoT platform using blockchain technology to guarantee sensing data integrity. The aim of this platform is to afford the device owner a practical application that provides a comprehensive, immutable log and allows easy access to their devices deployed in different domains. It also provides characteristics of general IoT systems, allows for real-time monitoring, and control between the end user and device. The business logic of the application is defined by the smart contract, which contains rules and conditions. The proposed approach is backed by a proof of concept implementation in realistic IoT scenarios, utilizing Raspberry Pi devices and a permissioned network called Hyperledger Fabric. Lastly, a benchmark study using various performance metrics is made to highlight the significance of the proposed work. The analysis results indicate that the designed platform is suitable for the resource-constrained IoT architecture and is scalable to be extended in various IoT scenarios.

The Do-It-Yourself (DIY) vision for the design of a smart and customizable IoT application demands the involvement of the general public in its development process. The general public lacks the technical knowledge for programming state-of-the-art prototyping and development kits. The latest IoT kits, for example, Raspberry Pi, are revolutionizing the DIY paradigm for IoT, and more than ever, a DIY intuitive programming interface is required to enable the masses to interact with and customize the behavior of remote IoT devices on the Internet. However, in most cases, these DIY toolkits store the resultant configuration data in local storage and, thus, cannot be accessed remotely. This paper presents the novel implementation of such a system, which not only enables the general public to customize the behavior of remote IoT devices through a visual interface, but also makes the configuration available everywhere and anytime by leveraging the power of cloud-based platforms. The interface enables the visualization of the resources exposed by remote embedded resources in the form of graphical virtual objects (VOs). These VOs are used to create the service design through simple operations like drag-and-drop and the setting of properties. The configuration created as a result is maintained as an XML document, which is ingested by the cloud platform, thus making it available to be used anywhere. We use the HTTP approach for the communication between the cloud and IoT toolbox and the cloud and real devices, but for communication between the toolbox and actual resources, CoAP is used. Finally, a smart home case study has been implemented and presented in order to assess the effectiveness of the proposed work.