Search

Surface quality inspection and control are extremely important for electronic manufacturing. The use of machine vision technology to automatically detect the defects of products has become an indispensable means for better quality control. A machine vision-based surface quality inspection system is usually composed of two processes: image acquisition and automatic defect detection. In this paper, we propose a deep learning-based approach for the defect detection of Copper Clad Laminate (CCL) images acquired from an industrial CCL production line. In the proposed approach, a new convolutional neural network (CNN) that realizes fast defect detection while maintaining high accuracy is designed. Our proposed approach makes four contributions. First, we introduce the depthwise separable convolution to reduce the calculation time. Second, we improve the squeeze-and-excitation block to improve network performance. Third, we introduce the squeeze-and-expand mechanism to reduce the computation cost. Fourth, we employ a smoother activation function (Mish) to allow improved information flow. The proposed network is compared with the benchmark CNNs (including Inception, ResNet and MobileNet). The experimental results show that compared with the benchmark networks, our proposed network has achieved the best results regarding the accuracy and suboptimal results in terms of the speed compared with the benchmark networks. Therefore, our proposed method has been integrated into an industrial CCL production line as a guideline for online defective product rejection.

In dense neighborhoods, there are often dozens of homes in close proximity. This can either be a tight city-block with many single-family homes (SFHs), or a multiple dwelling units (MDU) complex (such as a big apartment building or condominium). Each home in such a neighborhood (either a SFH or a single unit in a MDU complex) has its own Wi-Fi access point (AP). Because there are few (typically 2 or 3) non-overlapping radio channels for Wi-Fi, neighboring homes may find themselves sharing a channel and competing over airtime, which may cause bad experience of slow internet (long latency, buffering while streaming movies, etc.). Wi-Fi optimization over all the APs in a dense neighborhood is highly desired to provide the best user experience. We present a method for Wi-Fi channel selection in a centralized way for all the APs in a dense neighborhood. We describe how to use recent observations to estimate the potential-pain matrix - for each pair of APs, how much Wi-Fi-pain would they cause each other if they were on the same channel. We formulate an optimization problem - finding a channel allocation (which channel each home should use) that minimizes the total Wi-Fi-pain in the neighborhood. We design an optimization algorithm that uses gradient descent over a neural network to solve the optimization problem. We describe initial results from offline experiments comparing our optimization solver to an off-the-shelf mixed-integer-programming solver. In our experiments we show that the off-the-shelf solver manages to find a better (lower total pain) solution on the train data (from the recent days), but our neural-network solver generalizes better - it finds a solution that achieves lower total pain for the test data (tomorrow)., Comment: We discussed this work in the 2022 Fall Technical Forum as part of SCTE Cable-Tec Expo. This paper was published in SCTE Technical Journal. For citing this work, please cite the original publication

Additional file 1: Figure S1. The effect of cytochrome C concentrations on the SMA conversion rate. Figure S2. The cytochrome C catalyzed ATRP for the polymerization of PSS. The monomer of PSS do not show fluorescence (Ex: 485 nm; Em: 535 nm), while the polymerized PSS showed obvious fluorescence with the excitation of 485 nm. The results indicated that the PSS monomer could be polymerized by cytochrome C catalyzed ATRP under anaerobic or aerobic condition. Figure S3. GPC analysis of PSS polymer synthesized by the cytochrome C catalyzed ATRP. Figure S4. The time-course change of the Fe(II) concentration of the cytochrome C catalyzed ATRP under anaerobic condition. Figure S5. The effects of different metal ion chelators (2 mM) on the SMA conversion of the cytochrome C catalyzed ATRP under anaerobic condition. Figure S6. SMA conversion of the free Fe(II) or Fe-TPMA catalyzed ATRP under anaerobic or aerobic condition. Figure S7. The proposed schematic for cytochrome C catalyzed ATRP.

We report a highly stable humidity sensor based on nanocomposite film obtained by depositing Au nanoparticles on the surface of 3-mercaptopropionic acid (MPA) capped CdTe quantum dots (QDs) and the ...

While using a speaker verification (SV) based system in a commercial application, it is important that customers have an inclusive experience irrespective of their gender, age, or ethnicity. In this paper, we analyze the impact of gender and age on SV and find that for a desired common False Acceptance Rate (FAR) across different gender and age groups, the False Rejection Rate (FRR) is different for different gender and age groups. To optimize FRR for all users for a desired FAR, we propose a context (e.g. gender, age) adaptive thresholding framework for SV. The context can be available as prior information for many practical applications. We also propose a concatenated gender/age detection model to algorithmically derive the context in absence of such prior information. We experimentally show that our context-adaptive thresholding method is effective in building a more efficient inclusive SV system. Specifically, we show that we can reduce FRR for specific gender for a desired FAR on the voxceleb1 test set by using gender-specific thresholds. Similar analysis on OGI kids' speech corpus shows that by using an age-specific threshold, we can significantly reduce FRR for certain age groups for desired FAR.

We present a low-cost clinically viable ventilator design, AmbuBox, using a controllable pneumatic enclosure and standard manual resuscitators that are readily available (AmbuBag), which can be rapidly deployed during pandemic and mass-casualty events with a minimal set of components to manufacture and assemble. The AmbuBox is designed to address the existing challenges presented in the existing low-cost ventilator designs by offering an easy-to-install and simple-to-operate apparatus while maintaining a long lifespan with high-precision flow control. As an outcome, a mass-producible prototype of the AmbuBox has been devised, characterized, and validated in a bench test setup using a lung simulator. This prototype will be further investigated through clinical testing. Given the potentially urgent need for inexpensive and rapidly deployable ventilators globally, the overall design, operational principle, and device characterization of the AmbuBox system have been described in detail with open access online. Moreover, the fabrication and assembly methods have been incorporated to enable short-term producibility by a generic local manufacturing facility. In addition, a full list of all components used in the AmbuBox has been included to reflect its low-cost nature.

Aiming at the problem of extracting cracks on highway pavement under complex backgrounds such as water stains and rutting marks, the global image processing method is difficult to separate the foreground and the background, and the cracks cannot be extracted correctly. For this situation, this paper proposes a method based on probability statistics. The original image is divided into blocks, and then statistics and comparisons are performed on each image to complete the preliminary positioning of the crack area, and then perform the traditional image on the crack area. After treatment, the crack was finally extracted successfully. The experimental results show that the method proposed in this paper has a good effect and high accuracy for the extraction of pavement cracks with complex backgrounds.

Stable continuous micro-feeding of fine cohesive powders has recently gained importance in many fields. However, it remains a great challenge in practice because of the powder aggregate caused by interparticle cohesive forces in small capillaries. This paper describes a novel method of feeding fine cohesive powder actuated by a pulse inertia force and acoustic radiation force simultaneously in an ultrasonic standing wave field using a tapered glass nozzle. Nozzles with different outlet diameters are fabricated using glass via a heating process. A pulse inertia force is excited to drive powder movement to the outlet section of the nozzle in a consolidated columnar rod mode. An acoustic radiation force is generated to suspend the particles and make the rod break into large quantities of small agglomerates which impact each other randomly. So the aggregation phenomenon in the fluidization of cohesive powders can be eliminated. The suspended powder is discharged continuously from the nozzle orifice owing to the self-gravities and collisions between the inner particles. The micro-feeding rates can be controlled accurately and the minimum values for RespitoseSV003 and Granulac230 are 0.4 mg/s and 0.5 mg/s respectively. The relative standard deviations of all data points are below 0.12, which is considerably smaller than those of existing vibration feeders with small capillaries.

Classification is a pivotal function for many computer vision tasks such as object classification, detection, scene segmentation. Multinomial logistic regression with a single final layer of dense connections has become the ubiquitous technique for CNN-based classification. While these classifiers project a mapping between the input and a set of output category classes, they do not typically yield a comprehensive description of the category. In particular, when a CNN based image classifier correctly identifies the image of a Chimpanzee, its output does not clarify that Chimpanzee is a member of Primate, Mammal, Chordate families and a living thing. We propose a multilayer dense connectivity for concurrent prediction of category and its conceptual superclasses in hierarchical order by the same CNN. We experimentally demonstrate that our proposed network can simultaneously predict both the coarse superclasses and finer categories better than several existing algorithms in multiple datasets.

Information extraction from images is a key problem. In this paper, a fast, high-precision, multi-target image processing method based on machine vision is proposed to detect nanoparticles. The proposed methods include the following steps: image pre-processing, feature extraction, and result output. An auxiliary-Calibration plate is used as the auxiliary need to determine the measurement accuracy. Through the analysis and calculation of the characteristic scales of nanoparticles, our results are highly consistent with the experiment, indicating the accuracy of our method and algorithm. From the input test images to the output results, the algorithm only takes less than a second. For the image including more complexed structures, it can also be accurately detected. Furthermore, the algorithm is transferable and extensible, and can be applied in wider image processing fields.

A resonator with a right-angle conical reflector has been proposed to produce high-power CO2 laser beams. To analyze eigenfields of the right-angle conical reflector resonator, this paper adopts and demonstrates the transfer matrix method. In this paper, the mode-fields and corresponding losses are described as eigenvectors and eigenvalues of a transfer matrix according to the self-reproducing principle of laser field. By solving the transfer matrix for eigenvectors and eigenvalues, we obtain field distributions and losses of the dominant eigenmodes. The calculation results reveal that the right-angle conical reflector resonator could be used for a high-power CO2 laser to achieve low-order modes. However, the beam quality is reduced due to the residual blind-hole, which is in accord with the experimental result.

Coffee-ring structures, which are easy to prepare, have many hotspots for surface-enhanced Raman scattering (SERS). However, the inhomogeneous distribution of hotspots causes very poor SERS detection reproducibility, which strongly restricts their practical applications. In this paper, by combining a coffee-ring structure with the waveguide effect of optical fibers, we propose novel micro-coffee-ring-patterned fiber SERS probes fabricated by a laboratory-developed laser-induced dynamic dip-coating method. The mechanisms of micro-coffee-ring formation on fiber facet are analyzed in detail. With the help of a programmable dip-coater, micro-coffee-ring-patterned fiber SERS probes can be automatically and reproducibly fabricated, and they simultaneously show high sensitivity and good reproducibility in SERS detections. A detection limit lower than 10−8 M in aqueous solvent for thiram and methyl parathion (MP) is achieved, and the relative standard deviation (RSD) of the SERS peaks is less than 6%. In addition, benefiting from numerous hotspots provided by micro-coffee rings, this micro-coffee-ring-patterned fiber SERS probe presents superior performance in in-situ trace detection of target molecules in complex liquid environments. Detection sensitivities of 10−7 M for thiram in lake water and in orange juice and 10−6 M for melamine and tetracycline in liquid milk are achieved.

In recent decades, microfluidics in biological applications have experienced significant growth due to their advantages of small volume, low cost, short reaction time and high throughput. Almost all cross-section shapes of micro-channels in microfluidic chips are rectangular or triangular by the existing chip fabricating technologies, including hot embossing, lithography, etching and injection molding, etc. However, compared with the above micro-channel shapes, the circular one has the advantages in aspects of fluid flow, droplet generating, heat transfer and its replication for blood vessels. This paper presents a T-junction droplet microfluidic chip with circular cross-section micro-channels. The effect of micro-channel wettability, interfacial tension, velocity and flow rate of continuous phase on droplet size are simulated and mechanism of droplets generating process is explored. Comparing with continuous phase viscosity and interfacial tension, flow rate plays a decisive role in determining the droplet size which is in the range of 100–350 [Formula: see text]m according to the simulation result. The Capillary number is affected by the above three parameters and an estimating numerical method for generated droplet size was proposed according to the above simulation results and calculated by Capillary number. The droplets, the sizes of which were in the range of 20–400 [Formula: see text]m, were produced by varying the parameters of water and oil flow rates in the designed T-junction droplet microfluidic chip with circular cross-section micro-channels.

Tb 0.29 (Dy 1- x Pr x ) 0.71 Fe 1.97 ( x =0, 0.1, 0.2 and 0.3) alloys prepared by the directional solidification method were treated at 1073, 1123, 1173, 1223 and 1273 K for 4 h for homogenization, respectively. The magnetostriction, micro-morphology and composition distribution were studied by the standard resistance strain gauge technique, optical microscopy and scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDS). The results indicated that heat treatment could make the microstructure of alloys homogeneous effectively and improve the magnetostriction significantly. The optimum temperature was 1223 K. Compared to the as-cast ones, the increase amplitudes of magnetostriction of the corresponding samples were 39.5%, 64.9%, 95.3% and 50.8% when x =0, 0.1, 0.2 and 0.3, respectively at the compressive stress of 2 MPa and a magnetic field of 80 kA/m. The compressive stress could also improve the magnetostriction. However, further Pr element addition and over high heat treatment temperature would lead to the excessive decomposition of PrFe 2 and destroy the homogeneity, resulting in the decline of magnetostriction of alloys.

In microfluidic chips, most micro-channel cross-section shapes are rectangular or triangular in existing chip fabricating technologies, including hot embossing, lithography, etching, injection molding, etc. However, compared with the above micro-channel shapes, a circular shape has advantages in aspects of fluid flow, droplet generation, heat transfer and blood vessel replication. This paper presents a sandwich-like microfluidic chip with circular cross-section micro-channels. The sandwich-like structure includes three layers. The top and bottom layers are PDMS material while the middle layer is composed of micron glass capillaries (used as micro-channels) with circular cross-sections. The glass capillaries are made of borosilicate tube by a glass heating process. Sphere shaped paraffin wax is used as a sacrificial material to form micro-channel junctions. To test the functions of the fabricated micro sandwich-like microfluidic chip, a droplet generating experiment was conducted in the T-junction chip. The droplet size can be controlled in the range of 20–400 [Formula: see text]m by varying the water and oil flow rates. This proposed microfluidic chip structure has the advantages of short processing cycle, low cost and small flow resistance.

Micro-dosing of fine cohesive powders is the key technology in additive manufacturing and especially in high-potency active pharmaceutical ingredients (HPAPI). However, high accuracy micro-dosing (

A simulated annealing method for material parameter extraction with terahertz time-domain spectroscopy is introduced to improve the common extraction method. The simulated annealing method seeks the global minimum of the error function to obtain material constants where the characterized material is not constrained by a certain boundary condition. It is shown from the research results that more accurate material parameters can be acquired to meet the actual requirements of material analysis by use of the simulated annealing method.

High performance grid-controlled electron gun is the base of modern high power and broadband traveling-wave-tubes. We have developed a shadow grid electron gun, its perveance can be more than 1.6 micro-perveance. The parameters and diagram of the gun is given in this article. Better tactical performance, better anti-interference performance, good mobility are the important conditions for radar to alive. High power traveling-wave-tubes are the key device of the radar; they need high output power, flexible modulation method, and small permanent magnetic focusing system. Accordingly the electron gun of high power TWT should provide high perveance electron beam and should be shadow grid controlled. This article introduces a shadow grid controlled electron gun, the cathode voltage is 26kV, and the perveance can be up to 1.6 micro perveance. The calculated electron trajectories by the 2D compute program are shown. The affect of the circular wires of the shadow grid and the control grid to the moving of the electrons are taken into account during calculation. The shielding factor due to radial grid wires has also been considered.

Two new triterpene esters, named tereticornate A and B were isolated from the dried leaves of Eucalyptus tereticornis . The structures of these este

Two new C 17 -polyacetylenes and a C 14 -polyacetylene were isolated from dried roots of Panax quinquefolium . Their structures were determined by their 1 H NMR, 13 C NMR and mass spectral data. The cytotoxic activity of C 17 -polyacetylenes against leukemia cells (L 1210) was ca 20 times that of the C 14 -polyacetylene.