In recent years, deep learning has been successfully applied to hyperspectral image classification (HSI) problems, with several convolutional neural network (CNN) based models achieving an appealing classification performance. However, due to the multi-band nature and the data redundancy of the hyperspectral data, the CNN model underperforms in such a continuous data domain. Thus, in this article, we propose an end-to-end transformer model entitled SAT Net that is appropriate for HSI classification and relies on the self-attention mechanism. The proposed model uses the spectral attention mechanism and the self-attention mechanism to extract the spectral–spatial features of the HSI image, respectively. Initially, the original HSI data are remapped into multiple vectors containing a series of planar 2D patches after passing through the spectral attention module. On each vector, we perform linear transformation compression to obtain the sequence vector length. During this process, we add the position–coding vector and the learnable–embedding vector to manage capturing the continuous spectrum relationship in the HSI at a long distance. Then, we employ several multiple multi-head self-attention modules to extract the image features and complete the proposed network with a residual network structure to solve the gradient dispersion and over-fitting problems. Finally, we employ a multilayer perceptron for the HSI classification. We evaluate SAT Net on three publicly available hyperspectral datasets and challenge our classification performance against five current classification methods employing several metrics, i.e., overall and average classification accuracy and Kappa coefficient. Our trials demonstrate that SAT Net attains a competitive classification highlighting that a Self-Attention Transformer network and is appealing for HSI classification.
Despite significant progress in object detection tasks, remote sensing image target detection is still challenging owing to complex backgrounds, large differences in target sizes, and uneven distribution of rotating objects. In this study, we consider model accuracy, inference speed, and detection of objects at any angle. We also propose a RepVGG-YOLO network using an improved RepVGG model as the backbone feature extraction network, which performs the initial feature extraction from the input image and considers network training accuracy and inference speed. We use an improved feature pyramid network (FPN) and path aggregation network (PANet) to reprocess feature output by the backbone network. The FPN and PANet module integrates feature maps of different layers, combines context information on multiple scales, accumulates multiple features, and strengthens feature information extraction. Finally, to maximize the detection accuracy of objects of all sizes, we use four target detection scales at the network output to enhance feature extraction from small remote sensing target pixels. To solve the angle problem of any object, we improved the loss function for classification using circular smooth label technology, turning the angle regression problem into a classification problem, and increasing the detection accuracy of objects at any angle. We conducted experiments on two public datasets, DOTA and HRSC2016. Our results show the proposed method performs better than previous methods.
In the method of monitoring the liquid level based on ultrasonic impedance, the near-field effect can seriously affect the validity of the results. In this paper, we explore the factors affecting the length of the ultrasonic near field. Based on that, we propose the optimal length and the minimum length of the buffer block to avoid the near field. The evaluations show that when the parameters of the ultrasonic probe are 15 mm in diameter, 1 MHz in frequency, and ±15 V in emitted ultrasonic wave amplitude, the best results are obtained when the length of the buffer block is 22 mm. When the probe diameter is 10 mm, the buffer block length should be ≥5 mm to ensure the validity of the measured results. The evaluation precision is 1 mm. This research can effectively avoid the blind area of emitted waves when using ultrasonic to measure the liquid level. It provides an effective basis for the selection and design of ultrasonic probes.
Topology architecture has a decisive influence on network reliability. In this paper, we design a novel redundancy topology and analyze the structural robustness, the number of redundant paths between two terminal nodes, and the reliability of the proposed topology by using natural connectivity and time-independent and time-dependent terminal pair reliability, k-terminal reliability, and all-terminal reliability comprehensively and quantitatively, and we compare these measures of the proposed topology with AFDX in three scenarios. The evaluations show that in the structural robustness analysis, when no nodes are removed, the natural connectivity of the proposed topology with 10 nodes, 16 nodes, and 20 nodes is 77.8%, 26.95%, and 81.39% higher than that of AFDX, respectively. In the time-independent reliability analysis, when the link reliability is 0.9, terminal pair reliability of the proposed topology with 10 nodes, 16 nodes, and 20 nodes is 5.78%, 17.75%, and 34.65% higher than that of AFDX, respectively; k-terminal reliability is 10.04%, 31.97%, and 53.74% higher than that of AFDX, respectively; and all-terminal reliability is 29.36%, 74.37%, and 107.91% higher than that of AFDX, respectively. In the time-dependent reliability analysis, when the operating time is 8000 h, the terminal pair reliability of the proposed topology with 10 nodes, 16 nodes, and 20 nodes is 3.53%, 10.87%, and 21.08% higher than that of AFDX, respectively; the k-terminal reliability is 6.20%, 19.65%, and 32.58% higher than that of AFDX, respectively; and the all-terminal reliability is 18.25%, 45.04%, and 63.86% higher than that of AFDX, respectively. The proposed topology increases the redundant paths of data transmission. It ensures reliable data transmission and has high robustness and reliability. It provides a new idea for improving the reliability of industrial buses.
The ultrasonic Lamb wave detection principle can realize the noncontact measurement of liquid level in closed containers. When designing an ultrasonic Lamb wave sensor, it is vital to thoroughly study and select the optimal wedge size at the front of the sensor. In this paper, firstly, we select the best working mode of Lamb waves according to their propagation dispersion curve in aluminum alloy, and we obtain the best angle of wedge through experiments. Secondly, we study the impact of the size of the wedge block on the results, and we obtain the selection method of wedge block parameters. The evaluations show that, when the frequency–thickness product is 3 MHz·mm, the Lamb waves work in the A1 mode, and the experimental effect is the best. At this time, the incident angle of the ultrasonic wave is 27.39°. The wedge thickness should be designed to avoid the near-field area of the ultrasonic field, and we should choose the length as odd multiples of 1/4 wavelength. The rules obtained from the experiment can effectively select the best working mode for ultrasonic Lamb waves, while also providing a basis for the design of the wedge block size in a Lamb wave sensor.
The performance of the ultrasonic transducer will directly affect the accuracy of ultrasonic experimental measurement. Therefore, in order to meet the requirements of a wide band, a kind of annular 2-2-2 piezoelectric composite is proposed based on doped PDMS. In this paper, the transducer structure consisted of PZT-5A piezoelectric ceramics and PDMS doped with 3 wt.% Al2O3:SiO2 (1:6) powder, which constituted the piezoelectric composite. MATLAB and COMSOL software were used for simulation. Meanwhile, the electrode materials were selected. Then, the performance of the designed annular 2-2-2 ultrasonic transducer was tested. The simulation results show that when the polymer phase material of the piezoelectric ultrasonic transducer is doped PDMS, the piezoelectric phase and the ceramic substrate account for 70% of the total volume, the polymer phase accounts for 30% of the total volume, and the maximum frequency band width can reach 90 kHz. The experimental results show that the maximum bandwidth of −3 dB can reach 104 kHz when the frequency is 160 kHz. The results of the electrode test show that the use of Cu/Ti electrode improves the electrical conductivity of the single electrode. In this paper, the annular 2-2-2 transducer designed in the case of small volume had the characteristics of a wide frequency band, which was conducive to the miniaturization and integration of the transducer. Therefore, we believe that the annular 2-2-2 piezoelectric composite has broad application prospects.
Abstract. Focus on the problem that TI's dedicated video processor TMS320DM8148 cannot directly collect video data from HDMI interface, this paper presents a video coding system based on H.264.The system used DSP+FPGA architecture, FPGA is responsible for collecting video data of HDMI interface and caching. And then send to DM8148 through GPMC, DM8148 completes video encoding and decoding through the interaction of internal modules. The results show that video is displayed clearly and smoothly, without distortion or error. It successfully realizes the collection of the video data of HDMI interface obtained with DM8148.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.