Rapid development of affordable and portable consumer depth cameras facilitates the use information in many computer vision tasks such as intelligent vehicles 3D reconstruction. However, map captured by low-cost sensors (e.g., Kinect) usually suffers from low spatial resolution, which limits its potential applications. In this paper, we propose a novel deep network for super-resolution (SR), called DepthSR-Net. The proposed DepthSR-Net automatically infers high-resolution (HR) low-resolution...

During recent years, we have witnessed a rapid development of wireless network technologies which revolutionized the way people take and share multimedia content. However, images captured in outdoor scenes usually suffer from limited visibility due to suspended atmospheric particles, directly affects quality photos. Despite progress image dehazing methods, visual dehazed results still needs further improvement. In this paper, propose deep convolutional neural (CNN) for single called PDR-Net,...

The forest soil carbon pool plays a vital role in terrestrial ecosystems, being of great significance for maintaining global balance, regulating the cycle, and facilitating ecological restoration. Shandong Changyi Marine Ecological Special Protection Area is only state-level marine special protection area China with tamarisk as main object protection, it largest continuous best preserved natural distribution on mainland coast China. Compared to other forested areas, research spatial SOC at...

From the perspective of historical trajectory data and real-time data, a short-term coordinate point prediction method based on GRU (Gated Recurrent Unit) cyclic neural network is proposed. The main algorithm, in first stage, initialization training. parameters are learned by batch processing, which have been trained input to second stage as initial next network. trajectory. used values online adaptively update predicted flight points. Finally, compared with other typical time series models...

In this paper, a method to fast classify (Intradural hemorrhage, epidural and cerebral parenchymal hemorrhage) locate the bleeding points by using Singularity Expansion Method (SEM) Backpropagation (BP) neural network optimized genetic algorithm (GA) Sparrow Search Algorithm (SSA) is proposed. simulation model, spot with radius of 3 mm successfully identified approach. The test accuracy in for both bleeding's localization classification are 98.0% 97.4%, respectively. Head phantoms that have...

Received signal strength indicator (RSSI), the most accessible physical metric for commercial Ultra High Frequency Radio Identification (UHF RFID) systems, is highly affected by multi-path propagation, noise and other factors, making it difficult to achieve high ranging accuracy localization. Thus, similarity-based scene analysis algorithms become critical advantage of non-ranging. Nevertheless, as number reference tags used rises, tag antenna interference emerges a new crucial factor in...

In this paper, Adaboost and SVM are applied to SAR ATR (synthetic aperture radar automatic target recognition) respectively. The performance of these two classifiers is analyzed compared in aspect window with different size. First, PCA (principal component analysis) features selected as feature, then Adaboost.Ml used classify, Experimental results based on MSTAR data sets show that classifier has better robustness than

Aspect-based sentiment analysis aims to predict polarity for every aspect in a sentence review. Most existing approaches are based on the sequence models, which may superimpose emotional semantics of different tendencies and lack syntactic structure information. And most models adopt coarse-grained attention mechanism still face issues weakness interaction between context. In this paper, we propose transformer multi-grained network (T-MGAN), utilizes Transformer module learn word-level...

A novel runway detection algorithm for PolSAR (Polarimetric Synthetic Aperture Radar) images based on optimized polarimetric features and local spatial information is proposed. Existing methods always utilize the parallel line as primary feature. However, many other ground objects such rivers roads also have structures thus affect performance of these methods. The proposed method two stages classification with HOG (Histogram Oriented Gradient) feature, while avoiding interference due to...

To improve feature learning ability and accurately diagnose the faults of rolling bearings under a strong background noise environment, we present new shrinkage function named leaky thresholding to replace soft in deep residual networks (DRSNs). In this work, discover that such improved (IDRSNs) can be realized by using group searching method optimize slope value thresholding, IDRSNs more effectively eliminate signal features. We highlight our techniques significantly performance on various...

Abstract Recently, the fault diagnosis domain has witnessed a surge in popularity of deep residual shrinkage network (DRSN) due to its robust denoising capabilities. In our previous research, an enhanced version DRSN named global multi-attention (GMA-DRSN) is introduced augment feature extraction proficiency specifically for noised vibration signals. However, utilization multiple attention structures GMA-DRSN leads escalation computational complexity network, which may pose practical...

This paper presents an interactive and visual educational tool developed for teaching learning in the course of signals systems, digital signal processing. It is built with five main function modules, which are collecting, analysis, linear system filter design processing techniques application module. Matlab GUI used to tool. Feedback from students shows that this tool, they can understand knowledge points courses much easier their interest stimulated obviously. That means appropriate...

Automatic target recognition (ATR) based on synthetic aperture radar (SAR) imagery (denoted as SAR ATR for simplicity) is very important battlefield awareness. Since images are sensitive to pose variation of targets, a well-known challenging problem. An efficient algorithm given, which uses KFD (kernel Fisher discriminant) feature extractor and linear SVM (support vector machine) classifier. Experimental results evaluated with the MSTAR (moving stationary automatic recognition) public data...

With the development of remote sensing systems, scale imaging data grows rapidly, which highly requires appropriate adaptability for interpretation algorithms. Focusing on this trend, an unsupervised classification algorithm polarimetric synthetic aperture radar (PolSAR) images is proposed based multi-level feature extraction. The firstly generates initial map by extraction, and then introduces Wishart classifier into iterative to refine initial. At first level, PolSAR image classified four...

A novel short-term four-dimensional (4D) trajectory prediction model based on deep learning is proposed in this paper. The LSTM (Long Short-Term Memory) neural network. It consists of input layer, hidden layer and output layer. Original data first preprocessed order to form supervised sequences which are used as the model. cell information flow from each unit next moment includes state state, can be implicitly motion aircraft trajectory. Four-dimensional predicted obtained Experimental...

A novel signal processing method to directly extract the hemorrhage response signals using maximum correlation kurtosis deconvolution (MCKD) technique and complete ensemble empirical mode decomposition (CEEMDAN) algorithm is proposed in this article. This only needs acquire raw data measured one time, which saves a lot of time acquisition. In emergency cases, shorter detection means better prognosis life for patients. portable microwave brain imaging system comprising vector network analyzer...

When the extended Kalman filter (EKF) is applied in aircraft attitude estimation, two defects exist: one computational complexity; other large linearization error. Aiming at these problems, central difference (CDKF) based on Stirling interpolation formulation to low-cost estimation system which of less accurate and high noisy sensors. First, nonlinear mathematical model quaternion established, then CDKF estimation. Experimental results with real flying data show that superior commonly used...

Conventional sparsity-driven synthetic aperture radar (SAR) imagery often encounters the problem of loss structural features in weak scattering. Although there are algorithms that focus on structure enhancement, no proper balance between accuracy and efficiency can be achieved. In this article, a novel feature enhancement algorithm, named structure-awareness SAR (SA-SAR), is proposed by exploiting an emerging regularizer tensor total variation (STV). By imposing STV norm onto prior scenes or...

Spark has become a very attractive platform for big data analytics in recent years due to its unique advantages such as parallelism, fault tolerance, and complexity associated with clusters setup. On the spark platform, users can adjust parameter configurations according different job requirements specific applications optimize performance. This leads problem that we can't ignore, already more than 180 parameters, huge combination of parameters means rely on manual tuning grasp impact all In...

Choosing the right parameter configurations for recurring jobs running on big data analytics platforms is difficult because there can be hundreds of possible to pick from. Even selection based different types applications and user requirements. The difference between best configuration worst have a performance impact more than 10 times. However, parameters are not independent, which makes it challenge automatically identify optimal broad spectrum applications. To alleviate these problems, we...