Multicopters are attracting more and attention these years. In the design stage, designers users wonder if an assembled multicopter can meet their performance requirements, such as hovering endurance, system efficiency, maximum load, pitch, flight distance. However, in practice, they used to evaluate of a through lots experiments or by experience, which normally inefficient costly. This motivates us propose comprehensive offline evaluation algorithm performance. The indices considered mainly...

This paper proposes to utilize supervised deep convolutional neural networks take full advantage of the long-term spatial-temporal information in order improve video saliency detection performance. The conventional methods, which use temporally neighbored frames solely, could easily encounter transient failure cases when clues are less-trustworthy for a long period. To tackle aforementioned limitation, we plan identify those beyond-scope with trustworthy first and then align it current...

Path planning is the key for unmanned aerial vehicle (UAV) to perform tasks efficiently, which needs quickly obtain optimal path in complex environment. To solve problem urban environment, an improved artificial bee colony algorithm based on multi-strategy synthesis (IABC) proposed generate appropriate UAV. The IABC hybrid mechanism of chaotic mapping and Pareto principle initialize population, so as fully search solution space provide approximate flight Meanwhile, order balance exploration...

Driving patterns exert an important influence on the fuel economy of vehicles, especially hybrid electric vehicles. This paper aims to build a method identify driving with enough accuracy and less sampling time compared than other pattern recognition algorithms. Firstly identifier based Learning Vector Quantization neural network is established analyze six selected representative standard cycles. Micro-trip extraction Principal Component Analysis methods are applied ensure magnitude...

A major bottleneck of pedestrian detection lies on the sharp performance deterioration in presence small-size pedestrians that are relatively far from camera. Motivated by observation disparate spatial scales exhibit distinct visual appearances, we propose this paper an active detector explicitly operates over multiple-layer neuronal representations input still image. More specifically, convolutional neural nets, such as ResNet and faster R-CNNs, exploited to provide a rich discriminative...

The echo scattered from a slow-moving weak target on sea surface is nonstationary due to the influence of waves. Time-frequency distributions are good tools analyze it. In this paper, we propose method for detecting targets in clutter, which based time-frequency iteration decomposition. This consists three stages. First, present fast signal synthesis (FSSM) eigenvalue FSSM can synthesize faster and more accurately Wigner distribution (WD). Then, decomposition (IDM) masked WD FSSM. By IDM,...

Linear discriminant analysis (LDA) as a well-known supervised dimensionality reduction method has been widely applied in many fields. However, the lack of sparsity LDA solution makes interpretation results challenging. In this paper, we propose new model for sparse uncorrelated (ULDA). Our is based on characterization all solutions generalized ULDA. We incorporate into ULDA transformation by seeking with minimum ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML"...

We focus on the challenging problem of filamentary structure segmentation in both 2D and 3D images, including retinal vessels neurons, among others. Despite increasing amount efforts learning based methods to tackle this problem, there still lack proper data-driven feature construction mechanisms sufficiently encode contextual labelling information, which might hinder performance. This observation prompts us propose a approach learn structured features paper. The aim integrate local spatial...

This Digital image correlation method is a non-contact measurement that achieves deformation by tracking speckle patterns cover the measured surface and deform with surface. The quality of has significant direct impact on accuracy DIC. A global evaluation for pattern using simple easy-to-calculate parameter called two-dimensional entropy proposed in this paper. To verify effectiveness method, comparative experiments are conducted displacement three types speckle. results indicate larger...

A highly quantum-efficient green phosphor was developed by constructing energy transfer from Eu 2+ to Tb 3+ ions.

The micro‐arc oxidation (MAO) is widely used to improve the corrosion resistance of aluminum alloys by forming an in situ ceramic coating alumina. However, defects such as pores, cracks, and cavities are inevitable forms during fabrication, significantly reducing limiting further application MAO coatings. Addressing this issue requires a thorough understanding defect formation mechanisms before optimizing preparation parameters. In view this, underlying discharge channels, film growth,...

The differential evolution algorithm (DE) is one of the most powerful stochastic real-parameter optimization algorithms. theoretical studies on DE have gradually attracted attention more and researchers. However, few researches been done to deal with convergence conditions for DE. In this paper, a sufficient condition corollary global optima are derived by using infinite product. A framework satisfying then established. It also proved that two common mutation operators satisfy framework....

In view of the high cost and sparse spatial resolution offshore meteorological observations, ocean winds retrieved from satellites are valuable in wind resource assessment as a supplement to situ measurements. This study examines satellite synthetic aperture radar (SAR) images ENVISAT advanced SAR (ASAR) for mapping resources with resolution. Around 181 collected pairs data maps 13 stations Hangzhou Bay compared. The statistical results comparing speed SAR-based show standard deviation (SD)...

This paper aims to tackle the practically very challenging problem of efficient and accurate hand pose estimation from single depth images. A dedicated two-step regression forest pipeline is proposed: given an input image, step one involves mainly 3D location in-plane rotation using a pixel-wise forest. utilized in two which delivers final by similar model based on entire image patch. Moreover, our guided internally executing kinematic chain model. For unseen test parameters are estimated...

Several or even dozens of times spatial scale variation is one the major bottleneck for pedestrian detection. Although Region-based Convolutional Neural Network (R-CNN) families have shown promising results object detection, they are still limited to detect pedestrians with large variations due fixed receptive field sizes on a single convolutional output layer. In contrast previous methods that simply combined predictions feature maps different resolution, we propose scale-aware hierarchical...