To tackle the difficulties in detection and removal of impulse noise faced by existing filters, to further improve denoising performance, we propose an adaptive sequentially weighted median filter for image corrupted noise. In proposed method, a detector employing 3σ principle normal distribution local intensity statistics, is proposed; with neighborhood size, removal, which operator derived reference spatial distances from central noisy pixel, i.e., weighting coefficients are inversely...

The recently proposed robust principle component analysis (RPCA) has been successfully applied in background subtraction. However, low-rank decomposition makes sense on the condition that foreground pixels (sparsity patterns) are uniformly located at scene, which is not realistic real-world applications. To overcome this limitation, we reconstruct input video frames and aim to make only sparse space but also time. Therefore, propose a joint frame set division RPCA-based method for In...

To overcome the drawbacks of existing filters for salt and pepper noises, an adaptive probability filter is proposed. For image, it detects noises based on characteristic minimum maximum intensity values images, as well distribution noise. If noise‐free intensities in neighbourhood repeat with a certain probability, highest repeated frequency used to remove noise statistical significance; otherwise, median pixels Experiments show that proposed method capable detecting more accurately perform...

ABSTRACT Magnetic resonance imaging (MRI) offers comprehensive information about brain structures, enabling excellent performance in tumor segmentation using multimodal MRI many methods. Nonetheless, missing modalities are common clinical practice, which can significantly degrade performance. Current methods often struggle to maintain feature consistency and robustness fusion when face difficulties accurately capturing boundaries. In this study, we propose an adaptive edge‐oriented...

ABSTRACT Motor learning involves complex interactions between the cognitive and sensorimotor systems, which are susceptible to different levels of task load. While mechanism underlying load‐dependent regulations in functions has been extensively investigated, their influence on downstream execution motor skill remains less understood. The current study extends understanding whether how alters activation pattern by a longitudinal functional near‐infrared spectroscopy (fNIRS) 30 healthy...

ABSTRACT Precise segmentation of pneumonia lesions using deep learning has been a research focus in medical image segmentation, which convolutional neural networks (CNNs) excel at capturing local features through layers but struggle with global information, while Transformers handle and long‐range dependencies well require substantial computational resources data. Motivated by the recently introduced Mamba that effectively models less complexity, we develop novel network architecture order...

The stereo-match to find the corresponding points in left and right images from an optical tracking system is a crucial step 3-D point reconstruction. Unfortunately, presence of ghost markers remains issue which caused by indistinct correspondence when several are coplanar with two centers. Such can lead erroneous medical instruments patients. Some methods have been proposed for this issue, but they rarely applied systems. To address we propose robust method combining maximum silhouette...

This study investigated the complexity of electromyography (EMG) lower limb muscles when performing obstacle crossing tasks at different heights in poststroke subjects versus healthy controls. Five and eight controls were recruited to perform various (randomly set 10, 20, 30% leg's length). EMG signals recorded from bilateral biceps femoris (BF), rectus (RF), medial gastrocnemius, tibialis anterior during task. The fuzzy approximate entropy (fApEn) approach was used analyze signals. fApEn...

A method based on S transforms and support vector machines was presented for fault diagnosis of power electronics circuits in which the transform time-frequency analysis signal is used to extract features corresponding various faults. Then, types are identified through pattern recognition classifier, SVM. The simulation results show that proposed can accurately diagnose faults locate elements circuits. It also has excellent performance noise robustness calculation complexity, thus, having...

Due to the limitation of existing filters in detection and removal fixed value impulse noise, authors propose an iterative grouping median filter (IGMF) according characteristics noise intensity distribution. It sorts noise‐free pixels neighbourhood by intensity, divides sorted into groups depending on differences adjacent pixels, finally takes maximum group as noisy pixel. This strategy is performed iteratively full advantage previous denoising results. Experiments show that IGMF...

Patient-to-image registration is required for image-guided surgical navigation, but marker-based time consuming and subject to manual error. Markerless an alternative solution avoid these issues.This study designs a calibration board proposes geometric method calibrate the near-infrared tracking structured light components of proposed optical navigation system simultaneously.A planar cylinder are used evaluate accuracy calibration. The mean error experiment 0.035 mm, diameter 0.119 mm. A...

Abstract MRI and CT images have been routinely used in clinical practice for treatment planning of the head‐and‐neck (HAN) radiotherapy. Delineating organs‐at‐risk (OAR) is an essential step radiotherapy, however, it time‐consuming prone to inter‐observer variation. The existing automatic segmentation approaches are either limited by image registration or lack global spatial awareness, thus under‐performed when dealing with complex anatomies. Herein, we propose a full‐scale attention network...

Eye tracking is a valuable topic in computer vision. Appearance-based eye promising research direction recent years. Convolutional neural networks (CNN) had been used gaze estimation, which cover the significant variability appearance caused by unconstrained head motion. With computation capability of consumer devices rapidly evolving, accurate and efficient appearance-based has potential for multipurpose applications. Person-independent have limit improving estimation accuracy....

This paper presents a novel method for fingerprint image quality. Five features are extracted from the to analyze quality and feature vector is formed five features. Then SVM classifier, which can solve small-sample learning problems with good generalization, trained classify image. The separated into one of three classes, good-quality,medium-quality, or poor-quality. Experimental results on FVC2004 private database show that proposed an effective efficient scheme measure Our overcomes...

Predicting gaze point on mobile devices without calibration in unconstrained environments has great significance human computer interaction. Appearance-based estimation methods have been improved due to the recent advance convolutional neural network (CNN) models and availability of large-scale datasets. CNN limitations extracting global information features ignore important local features. In this paper, we propose a novel structure named GazeAttentionNet. To improve accuracy estimation,...

Abstract Colonoscopy is one of the most direct and effective methods for detecting colon polyps; they are crucial early screening prevention colorectal cancer (CRC). Accurate segmentation polyp images significant clinical management treatment CRC. However, image a challenging task because polyps vary in size, shape, color, have low contrast with surrounding tissue mucosa. To address these challenges, we propose novel network called BGNet segmentation. consists three modules: boundary feature...