<abstract> <p>Medical image segmentation of the liver is an important prerequisite for clinical diagnosis and evaluation cancer. For automatic from Computed Tomography (CT) images, we proposed a Multi-scale Feature Extraction Enhancement U-Net (mfeeU-Net), incorporating Res2Net blocks, Squeeze-and-Excitation (SE) Edge Attention (EA) blocks. The blocks which are conducive to extracting multi-scale features were used as backbone encoder, while SE also added encoder enhance channel...

The challenge of allowing a quadrotor to land quickly on an unknown moving platform using visual cues is addressed. A position-based servoing (PBVS) framework designed that utilises the relative position data captured by onboard camera ensure successful landing. inherent limitation imposed camera's low sampling rate, which hampers update and transfer rate control commands, mitigated introducing alternating predictive observer (APO). This inputs rapid actual or virtual information into...

Accurate segmentation of liver tumor regions in medical images is great significance for clinical diagnosis and the planning surgical treatments. Recent advancements machine learning have shown that convolutional neural networks are powerful such image processing while largely reducing human labor. However, variable shape, fuzzy boundary, discontinuous region tumors bring challenges to accurate segmentation. The feature extraction capability a network can be improved by expanding its...

Dual-rate dynamic systems consisting of a sensor with relatively slow sampling rate and controller/actuator fast updating widely exist in control systems. The bandwidth these dual-rate is severely restricted by the sensors, resulting various issues like sluggish dynamics closed-loop systems, poor robustness performance. A novel alternating predictive observer (APO) proposed to significantly enhance generic Specifically, at each period, we will first implement <italic...

This paper investigates the problem of robust output feedback control tracking for a class nonlinear dual-rate sample-data systems subject to time-varying disturbances. These systems, known as slow measurement fast (SMFC) are characterised by sampling rate sensor and updating controller. To facilitate high-frequency signal tracking, this introduces an alternating predictive generalised proportional integral observer (APGPIO). The proposed method not only effectively restores system state...

Abstract Segmentation of a tumor region from medical images is critical for clinical diagnosis and the planning surgical treatments. Recent advancements in machine learning have shown that convolutional neural networks are powerful such image processing while largely reducing human labor. However, variant shapes liver tumors with blurred boundaries cause great challenge accurate segmentation. The feature extraction capability network can be improved by expanding its architecture, but it...

In this paper, the anti-disturbance tracking control arithmetic for a class of MIMO nonlinear systems with input saturation is discussed. order to better characterize unknown disturbances, exogenous neural network adjustable parameters are employed and disturbance-observer-based-control (DOBC) framework also established appropriate systematic rules. Via using polytopic representation function, combining state feedback estimates disturbance, composite controller designed guarantee error...

Segmentation of tumor regions from medical images is crucial for clinical diagnosis. U-Net a practical encoder-decoder framework proposed better image segmentation. In segmentation image, performs more excellent than other network. At present, usually continuously deepens the network to improve feature extraction capability accuracy segmentation, rarely considering performance This paper aims simplify layer, segment liver area by mobilenetv3. Then, attention mechanism added skip connection...

This paper investigates the combination of output feedback anti-disturbance control with model predictive for slow-rate sampled-data systems non-vanishing disturbances. The study focuses on a specific class dual-rate systems, where sensor samples data at slow rate, and controller updates input fast rate. To address challenge sampling, introduces an observation method suitable alternating observer, utilizing only measured output. By matching observer's rate rapidly updated controller's...

This paper discusses the novel anti-disturbance control algorithm for hypersonic flight vehicle (HFV) models by using neural network (NN) identifier. Different from those existed results, unknown exogenous disturbances in HFV are assumed to be described designed NNs with adjustable parameters. Furthermore, disturbance-observer-based-control (DOBC) adaptive regulation laws is thus presented estimate nonlinear disturbances. By integrating estimated value of PI feedback input, a composite...