Aerial object tracking has recently shown great potential in the field of remote sensing. However, small objects with limited feature information pose a huge challenge to aerial trackers. Despite significant improvements, most trackers still struggle capture enough discriminative features and overcome background disturbances. In this work, we propose an efficient tracker (SmallTrack) based on Siamese network improve discrimination objects. It consists two effective modules, namely Wavelet...

This paper addresses the problem of visual object tracking for Unmanned Aerial Vehicles (UAVs). Most Siamese trackers are used to regard as classification and regression problems. However, it is difficult these accurately classify in face similar objects, background clutters other common challenges UAV scenes. So, a reliable classifier key improving performance. In this paper, simple yet efficient tracker following basic architecture neural network proposed, which improves ability from three...

The storage period of tea is a major factor affecting quality. However, the effect years on non-volatile functional components and quality green remains largely unknown. In this study, comparative analysis organic teas with varying (1–16 years) was conducted by quantifying 47 components, using electronic tongue chromatic aberration technology, alongside an evaluation antioxidative capacity. results indicated significant negative correlation between levels polyphenols, total amino acids,...

The creation of heterostructures with inherent interface polarization has been proven effective in enhancing piezocatalytic activity; however, developing efficient heterostructure piezocatalysts remains challenging, and the underlying mechanisms are not well understood. In this work, a stable Bi2WO6/BiOBr strong chemical binding is successfully constructed by exchanging double Br- WO4 2- hydrothermal reaction. demonstrates exceptional hydrogen evolution reaction (HER) efficiencies 0.75 mmol...

Abstract Recently, Siamese‐based trackers have drawn amounts of attention in visual tracking field because their excellent performance. However, object on Unmanned Aerial Vehicles platform encounters difficulties under circumstances such as small objects and similar interference. Most existing methods for aerial adopt deeper networks or inefficient policies to promote performance, but most can hardly meet real‐time requirements mobile platforms with limited computing resources. Thus, this...

Accurate prediction of mesoscale eddy trajectories requires efficient models with large-size available data instances to capture the main characteristics. However, there is a lack salient attention mechanisms that can recognize demand extracting aggregated features over multi-dimensional sources. Additionally, deep learning techniques are very important for trajectory dynamically properties eddies in South China Sea (SCS). In this paper, we propose spatio-temporal attention-based framework...

Purpose This paper aims to investigate the relative translational control for multiple spacecraft formation flying. proposes an engineering-friendly, structurally simple, fast and model-free algorithm. Design/methodology/approach a tanh-type self-learning (SLC) approach with variable learning intensity (VLI) guarantee global convergence of tracking error. algorithm utilizes controller's previous information in addition current system state avoids complicating structure. Findings The proposed...

An infrared image vehicle detection algorithm based on Haar-like features is proposed in this paper. Firstly, the top-hat transformation and bottom-hat are used to enhance image, contrast between background improved. Secondly, we propose use describe calculate vehicles' features. Thirdly, improved maximum entropy segmentation segment vehicle. After that, using vehicle's prior knowledge remove detected false alarm target. Finally, Fl-measure model evaluate algorithm. The experimental results...

Quadrotor helicopters are emerging as a popular platform for unmanned aerial vehicle (UAV) research. This work included integral action in Backstepping design attitude stabilization control of quadrotor helicopter. Firstly, complete dynamic model was built using blade element and momentum theories. Based on this new controller called `Integral Backstepping' designed by added tracking error term to the normal controller. Through various simulations real flight experiment, it is proved that...

The target positioning algorithm of the traditional unmanned aerial vehicle (UAV) airborne optoelectronic platform introduces a large number angle measurement errors, resulting in low accuracy. In this paper, hybrid nonlinear least squares and Gauss-Newton is proposed. Firstly, iterative localization based on laser ranging value derived. Then rough solution linear square used as initial Newton iteration method for location estimation. combines advantages simple easy implementation high...

This paper mainly introduces the research principle and implementation method of small UAV navigation attitude system based on MEMS devices. The Gauss - Newton least squares is used to calibrate accelerometer gyroscope for calibration. Improve accuracy by using modified complementary filtering correct angle error. experimental data show that design in this meet requirements achieve a small, low cost.

UAV flight safety is the first issue of UAV's attention. The establishment a monitoring system not only meets needs army but also has important significance. This paper introduces development status and trends aircraft fault prediction health management at home abroad. key technologies are analyzed, Including state parameters acquisition technology, parameter data real-time analysis condition model established establishment, security model, online control instruction technology assessment....

Abstract This article investigates the issue of neural network (NN)-based prescribed performance control with collision avoidance ability for a spacecraft formation system in presence space perturbations and thruster faults. First, an artificial potential function is constructed to guarantee that remain within communication range collision-free. Then, introduced such position errors are constrained preset boundary. Further, learning non-singular terminal sliding mode (LNTSMC) law explored...

This paper investigates the problem of attitude tracking in quadrotor unmanned aerial vehicles (UAVs) using a model-free online learning control (MFOLC) scheme. The system, which is represented by unit quaternions, considered presence uncertain and unknown inertia parameters, time-varying external disturbances, angular velocity measurement noise. A computationally low-cost scheme consisting baseline controller module capable from previous input designed. proposed does not require precise...