Feature Detection Research Articles

Principal feature detection via ϕ-Sobolev inequalities

Principal feature detection via ϕ-Sobolev inequalities

SEGMENTATION OF DIGITAL IMAGES WITH WAVELET TRANSFORMATION USING MATLAB VERSION R2010B

The purpose of this research is to explore and apply the use of wavelet transformation for the segmentation of digital images, utilizing MATLAB version R2010B. The study aims to analyze how wavelet transform can be used to enhance the accuracy and effectiveness of image segmentation, which is a critical process in image processing and computer vision. The research contributes to the field of digital image processing by demonstrating the application of wavelets transformation for segmenting digital images. In this study, the study provides insights into how wavelets can be utilized to improve the detection of image features, especially in identifying image features more accurately. The experimental results show that only the Canny operator's edge detection method has the best edge detector in detecting the edges of objects in wavelet images. The technique of determining the threshold value (thresholding) can be carried out in two ways, namely, automatic method and technique carried out by trial and error. Finally, improving automatic thresholding techniques using AI-driven algorithms to reduce the reliance on trial-and-error methods could provide more consistent results in varied application areas.

Feature detection of B-ultrasound images of intussusception in children based on improved YOLOv8n

To assist grassroots sonographers in accurately and rapidly detecting intussusception lesions from children's abdominal ultrasound images, this paper proposes an improved YOLOv8n children's intussusception detection algorithm, called EMC-YOLOv8n. Firstly, the EfficientViT network with a cascaded group attention module was used as the backbone network to enhance the speed of target detection. Secondly, the improved C2fMBC module was used to replace the C2f module in the neck network to reduce network complexity, and the coordinate attention (CA) module was introduced after each C2fMBC module to enhance attention to positional information. Finally, experiments were conducted on the self-built dataset of intussusception in children. The results showed that the recall rate, average detection accuracy (mAP@0.5) and precision of the EMC-YOLOv8n algorithm improved by 3.9%, 2.1% and 0.9%, respectively, compared to the baseline algorithm. Despite slightly increased network parameters and computational load, significant improvements in detection accuracy enable efficient completion of detection tasks, demonstrating substantial economic and social value.

Large-scale medieval urbanism traced by UAV-lidar in highland Central Asia.

Aerial light detection and ranging (lidar) has emerged as a powerful technology for mapping urban archaeological landscapes, especially where dense vegetation obscures site visibility1,2. More recently, uncrewed aerial vehicle/drone lidar scanning has markedly improved the resolution of three-dimensional point clouds, allowing for the detection of slight traces of structural features at centimetres of detail across large archaeological sites, a method particularly useful in areas such as mountains, where rapid deposition and erosion irregularly bury and expose archaeological remains3. Here we present the results of uncrewed aerial vehicle-lidar surveys in Central Asia, conducted at two recently discovered archaeological sites in southeastern Uzbekistan: Tashbulak and Tugunbulak. Situated at around 2,000-2,200 m above sea level, these sites illustrate a newly documented geography of large, high-altitude urban centres positioned along the mountainous crossroads of Asia's medieval Silk Routes (6th-11th centuryCE (Common Era)4,5. Although hidden by centuries of surface processes, our pairing of very-high-resolution surface modelling with semiautomated feature detection produces a detailed plan of monumental fortifications and architecture spanning 120 ha at Tugunbulak, thereby demonstrating one of the largest highland urban constellations in premodern Central Asia. Documentation of extensive urban infrastructure and technological production among medieval communities in Central Asia's mountains-a crucial nexus for Silk Road trade networks6-provides a new perspective on the participation of highland populations in the economic, political and social formation of medieval Eurasia.

Detection of strong spectral features from asymptotic-giant-branch stars in distant quiescent galaxies

Detection of strong spectral features from asymptotic-giant-branch stars in distant quiescent galaxies

Autonomous Positioning for Mobile Vehicles Based on Visual‐Inertial Fusion in Challenging Dark Roadway Scenes

ABSTRACTAccurate positioning for autonomous driven underground mining vehicles (UMVs) and coal mine robots (CMRs) is indeed one of the cores in the intelligentization of coal mining. Completely different from positioning on the ground and in parking scenes, there will be great difficulties in realizing the accurate active positioning for UMVs shuttled in dark and narrow roadways. We propose an effective visual and inertial fusion positioning method for autonomous CMRs and roadheaders in challenging roadway scenarios based on the odometer‐aided inertial navigation system and visual pose estimation system. Velocity information of the odometer is adapted to restrain the error accumulation of inertial positioning based on a Kalman filter. The hybrid visual feature detection algorithm is put forward to improve the accuracy and robustness of visual observation information in a dark environment. Autonomous experiments for CMRs and roadheaders are separately performed in the narrow roadway and dark passageway to demonstrate the applicability of our localization method. The proposed approach outperforms the subsystems and existing methods in accuracy and has outstanding stability.

PhoXplex: Combining Phospho-enrichable Cross-Linking with Isobaric Labeling for Quantitative Proteome-Wide Mapping of Protein Interfaces.

Integrating cross-linking mass spectrometry (XL-MS) into structural biology workflows provides valuable information about the spatial arrangement of amino acid stretches, which can guide elucidation of protein assembly architecture. Additionally, the combination of XL-MS with peptide quantitation techniques is a powerful approach to delineate protein interface dynamics across diverse conditions. While XL-MS is increasingly effective with isolated proteins or small complexes, its application to whole-cell samples poses technical challenges related to analysis depth and throughput. The use of enrichable cross-linkers has greatly improved the detectability of protein interfaces in a proteome-wide scale, facilitating global protein-protein interaction mapping. Therefore, bringing together enrichable cross-linking and multiplexed peptide quantification is an appealing approach to enable comparative characterization of structural attributes of proteins and protein interactions. Here, we combined phospho-enrichable cross-linking with TMT labeling to develop a streamline workflow (PhoXplex) for the detection of differential structural features across a panel of cell lines in a global scale. We achieved deep coverage with quantification of over 9000 cross-links and long loop-links in total including potentially novel interactions. Overlaying AlphaFold predictions and disorder protein annotations enables exploration of the quantitative cross-linking data set, to reveal possible associations between mutations and protein structures. Lastly, we discuss current shortcomings and perspectives for deep whole-cell profiling of protein interfaces at large-scale.

Design and Testing of a Seedling Pick-Up Device for a Facility Tomato Automatic Transplanting Machine.

At present, tomato transplanting in facility agriculture is mainly manual operation. In an attempt to resolve the problems of high labor intensity and low efficiency of manual operation, this paper designs a clip stem automatic transplanting and seedling picking device based on the yolov5 algorithm. First of all, through the study of the characteristics of tomato seedlings of different seedling ages, the age of tomato seedlings suitable for transplanting was obtained. Secondly, the improved yolov5 algorithm was used to determine the position and shape of tomato seedlings. By adding a lightweight upsampling operator (CARAFE) and an improved loss function, the feature extraction ability and detection speed of tomato seedling stems were improved. The accuracy of the improved yolov5 algorithm reached 92.6%, and mAP_0.5 reached 95.4%. Finally, the seedling verification test was carried out with tomato seedlings of about 40 days old. The test results show that the damage rate of the device is 7.2%, and the success rate is not less than 90.3%. This study can provide a reference for research into automatic transplanting machines.

Bilateral Defect Cutting Strategy for Sawn Timber Based on Artificial Intelligence Defect Detection Model.

Solid wood is renowned as a superior material for construction and furniture applications. However, characteristics such as dead knots, live knots, piths, and cracks are easily formed during timber's growth and processing stages. These features and defects significantly undermine the mechanical characteristics of sawn timber, rendering it unsuitable for specific applications. This study introduces BDCS-YOLO (Bilateral Defect Cutting Strategy based on You Only Look Once), an artificial intelligence bilateral sawing strategy to advance the automation of timber processing. Grounded on a dual-sided image acquisition platform, BDCS-YOLO achieves a commendable mean average feature detection precision of 0.94 when evaluated on a meticulously curated dataset comprising 450 images. Furthermore, a dual-side processing optimization module is deployed to enhance the accuracy of defect detection bounding boxes and establish refined processing coordinates. This innovative approach yields a notable 12.3% increase in the volume yield of sawn timber compared to present production, signifying a substantial leap toward efficiently utilizing solid wood resources in the lumber processing industry.

Core-Periphery Multi-Modality Feature Alignment for Zero-Shot Medical Image Analysis.

Multi-modality learning, exemplified by the language-image pair pre-trained CLIP model, has demonstrated remarkable performance in enhancing zero-shot capabilities and has gained significant attention recently. However, simply applying language-image pre-trained CLIP to medical image analysis encounters substantial domain shifts, resulting in severe performance degradation due to inherent disparities between natural (non-medical) and medical image characteristics. To address this challenge and uphold or even enhance CLIP's zero-shot capability in medical image analysis, we develop a novel approach, Core-Periphery feature alignment for CLIP (CP-CLIP), to model medical images and corresponding clinical text jointly. To achieve this, we design an auxiliary neural network whose structure is organized by the core-periphery (CP) principle. This auxiliary CP network not only aligns medical image and text features into a unified latent space more efficiently but also ensures alignment driven by principles of brain network organization. In this way, our approach effectively mitigates and further enhances CLIP's zero-shot performance in medical image analysis. More importantly, the proposed CP-CLIP exhibits excellent explanatory capability, enabling the automatic identification of critical disease-related regions in clinical analysis. Extensive experiments and evaluation across five public datasets covering different diseases underscore the superiority of our CP-CLIP in zero-shot medical image prediction and critical features detection, showing its promising utility in multimodal feature alignment in current medical applications.

MSFragger-DDA+ Enhances Peptide Identification Sensitivity with Full Isolation Window Search.

Liquid chromatography-mass spectrometry (LC-MS) based proteomics, particularly in the bottom-up approach, relies on the digestion of proteins into peptides for subsequent separation and analysis. The most prevalent method for identifying peptides from data-dependent acquisition (DDA) mass spectrometry data is database search. Traditional tools typically focus on identifying a single peptide per tandem mass spectrum (MS2), often neglecting the frequent occurrence of peptide co-fragmentations leading to chimeric spectra. Here, we introduce MSFragger-DDA+, a novel database search algorithm that enhances peptide identification by detecting co-fragmented peptides with high sensitivity and speed. Utilizing MSFragger's fragment ion indexing algorithm, MSFragger-DDA+ performs a comprehensive search within the full isolation window for each MS2, followed by robust feature detection, filtering, and rescoring procedures to refine search results. Evaluation against established tools across diverse datasets demonstrated that, integrated within the FragPipe computational platform, MSFragger-DDA+ significantly increases identification sensitivity while maintaining stringent false discovery rate (FDR) control. It is also uniquely suited for wide-window acquisition (WWA) data. MSFragger-DDA+ provides an efficient and accurate solution for peptide identification, enhancing the detection of low-abundance co-fragmented peptides. Coupled with the FragPipe platform, MSFragger-DDA+ enables more comprehensive and accurate analysis of proteomics data.

Lightweight image denoising method for feature-based visual SLAM in γ radiation environments

In the γ radiation environments, high-energy photons impact image sensor pixels and induce speckle noise in the scene image, which influences the accuracy of feature detection and matching in the visual Simultaneous Localization and Mapping (SLAM) algorithms. To mitigate these effects, this paper presents a lightweight image denoising method based on super-neighborhood down-sampling. Firstly, the super-neighborhood down-sampling strategy is used to split the speckle noise in the camera images into isolated point noises in the sub-image sets. Secondly, these isolated flickering point noises are removed by the detection-based median filter. Finally, the denoised scene image is reconstructed by the super-neighborhood up-sampling. Extensive experiments were performed on the radiation environment datasets that were captured by the camera inside the Cobalt-60 hot cell. The results show that the proposed method can effectively filter out the speckle radiation noise induced by γ rays, significantly improve the quality of features and feature matching rate, and improve the performance of ORB-SLAM3, the state-of-the-art feature-based SLAM algorithm, in radiation environments.

Enhancing Semantic Segmentation in High-Resolution TEM Images: A Comparative Study of Batch Normalization and Instance Normalization

Abstract Integrating deep learning into image analysis for transmission electron microscopy (TEM) holds significant promise for advancing materials science and nanotechnology. Deep learning is able to enhance image quality, to automate feature detection, and to accelerate data analysis, addressing the complex nature of TEM datasets. This capability is crucial for precise and efficient characterization of details on the nano—and microscale, e.g., facilitating more accurate and high-throughput analysis of nanoparticle structures. This study investigates the influence of batch normalization (BN) and instance normalization (IN) on the performance of deep learning models for semantic segmentation of high-resolution TEM images. Using U-Net and ResNet architectures, we trained models on two different datasets. Our results demonstrate that IN consistently outperforms BN, yielding higher Dice scores and Intersection over Union metrics. These findings underscore the necessity of selecting appropriate normalization methods to maximize the performance of deep learning models applied to TEM images.

Landmark-aware autonomous odometry correction and map pruning for planetary rovers

Planetary rover autonomous localization is paramount for a planetary surface exploration mission. However, existing methods demonstrate limited localization accuracy, mostly due to the unstructured texture characterization of planetary surface. In response, this study presents a novel Neural Radiance Field (NeRF) driven visual odometry correction method that allows for high-precision 6-DoF rover pose estimation and local map pruning. First, an innovative image saliency evaluation approach, combining binarization and feature detection, is introduced to meticulously select landmarks that are conducive to rover re-localization. Subsequently, we conduct 3D reconstruction and rendering of the chosen landmarks based on a-priori knowledge of planetary surface images and their Neural Radiance Field (NeRF) models. High-precision odometry correction is achieved through the optimization of photometric loss between NeRF rending images and real images. Simultaneously, the odometry correction mechanism is employed in an autonomous manner to refine the NeRF model of the corresponding landmark, leading to an improved local map and gradually enhanced rover localization accuracy. Numerical simulation and experiment trials are carried out to evaluate the performance of the proposed method, results of which demonstrate state-of-the-art rover re-localization accuracy and local map pruning.

High precision DSRC and LiDAR data integration positioning method for autonomous vehicles based on CNN

In order to improve the safe driving and automatic positioning capability of autonomous vehicles, a high-precision DSRC and LiDAR data integration positioning technology for autonomous vehicles based on CNN is proposed. Import the data of Dedicated Short Range Communications and Light Detection and Ranging for automatic driving vehicle positioning, carry out kinematic analysis of autonomous driving vehicles under multi-sensor fusion, and transform the data of DSRC and LiDAR sensors into tightly coupled coordinate systems; The CNN depth learning method is used to compensate the position and attitude tracking estimation error under the overall time stamp synchronization of the sensor through adaptive information tracking; The first and second order feedforward compensation is made for the positioning parameters of the autonomous driving vehicle using the PID model, and the point cloud feature matching model is fused to complete the estimation of the positioning attitude parameters of the autonomous driving vehicle. In order to eliminate the noise interference under the DSRC communication mechanism, the Kalman filter function is used to automatically optimize the constraint parameters in the point cloud feature detection model, and the positioning error parameters are dynamically filtered and adjusted; Kinematics analysis is carried out for the driving state of the vehicle, and the positioning error in the vehicle movement is controlled through the difference technology to achieve high-precision DSRC and LiDAR data integration positioning. The simulation results show that this method can integrate and locate the high-precision DSRC and LiDAR data of the autonomous vehicle, and the attitude estimation and positioning accuracy of the vehicle is good, while the error of the attitude parameter estimation of the autonomous vehicle is low.

A Novel Method for Identifying Landslide Surface Deformation via the Integrated YOLOX and Mask R-CNN Model

The detection of landslide areas and surface characteristics is the prerequisite and basis of landslide hazard risk assessment. The traditional method relies mainly on manual field identification, and discrimination is based on the lack of unified quantitative standards. Thus, the use of neural networks for the quantitative identification and prediction of landslide surface deformation is explored. By constructing an integrated model based on YOLO X-CNN and Mask R-CNN, a deep learning-based feature detection method for landslide surface images is proposed. First, the method superimposes Unmanned Aerial Vehicle (UAV) oblique photography data (UOPD) and Internet heterosource image data (IHID) to construct a landslide surface image dataset and landslide surface deformation database. Second, an integrated model suitable for small- and medium-scale target detection and large-scale target edge extraction is constructed to automatically identify and extract landslide surface features and to achieve rapid detection of landslide surface features and accurate segmentation and deformation recognition of landslide areas. The results show that the detection accuracy for small rock targets is greater than 80% and that the speed is 57.04 FPS. The classification and mask segmentation accuracies of large slope targets are approximately 90%. A speed of 7.89 FPS can meet the needs of disaster emergency response; this provides a reference method for the accurate identification of landslide surface features.

Photonic Neuromorphic Processing with On‐Chip Electrically‐Driven Microring Spiking Neuron

AbstractGuided by brain‐like temporal processing and event‐driven manner, neuromorphic computing has emerged as a competitive paradigm to realize artificial intelligence with high energy efficiency. Silicon photonics offers an ideal hardware platform with mutual foundry fabrication process and well‐developed device libraries, however, its huge potential to build integrated neuromorphic systems is significantly hindered due to the lack of scalable on‐chip photonic spiking neurons. Here, the first integrated electrically‐driven spiking neuron based on a silicon microring under the carrier injection working mode is reported, which is capable of emulating fundamental neural dynamics including excitability threshold, temporal integration, refractory period, controllable spike inhibition, and precise time encoding at a speed of 250 MHz. By programming time‐multiplexed spike representations, photonic spiking convolution is experimentally realized for image edge feature detection. Besides, a spiking convolutional neural network is constructed by combining photonic convolutional layers with a software‐implemented fully‐connected layer, which yields a classification accuracy of 94.1% on the benchmark Modified National Institute of Standards and Technology database. Moreover, it is theoretically verified that it's promising to further improve the operation speed to a gigahertz level by developing an electro‐optical co‐simulation model. The proposed microring neuron constitutes the final building block of scalable spike activation, thus representing a great breakthrough to boost the development of on‐chip neuromorphic information processing.

An Autonomous Feature Detection Method for Slow-Moving Small Target on Sea Surface Based on Kernelized Contextual Bandit

An Autonomous Feature Detection Method for Slow-Moving Small Target on Sea Surface Based on Kernelized Contextual Bandit

Fabrication of TPU-Supported Au-Bi2S3/Ti3C2Tx electrospun mat: Towards flexible and Multi-functional sensors for human motion and NH3 detection at room temperature

The potential applications of wearable sensors in health monitoring, micro-environment detection, and advanced soft electronic noses have attracted widespread attention. However, due to the inherent limitations of traditional organic flexible substrates, simultaneously achieving excellent flexibility, high sensitivity, robustness, and breathability remains a significant challenge. Herein, using the electrospinning method, a flexible TPU substrate film was prepared, and Bi2S3/Ti3C2Tx nanocomposite materials were synthesized through a simple one-step hydrothermal method. Au nanoparticles were attached to the surface of the composite material using sodium borohydride, and finally, the surface impregnation method was used to successfully prepare the Au-Bi2S3/Ti3C2Tx-TPU flexible electrospun mat. The as-prepared sensor demonstrated an outstanding response of 154 % to 100 ppm NH3 at room temperature, coupled with high selectivity, repeatability, and long-term stability. The strain properties of the sensor were also evaluated, showcasing its potential for human motion and physiological feature detection, such as breathing, swallowing, and limb movements. The study’s findings indicated that the Au-Bi2S3/Ti3C2Tx-TPU electrospun mat sensor has promising applications in flexible electronics, wearable environmental monitoring devices, and human healthy monitoring.

Dynamic digital image correlation method for rolling convective contact

Digital image correlation (DIC) is an increasingly popular and effective non-contact method for measuring full-field displacements and strains of deformable bodies under load. Current DIC methods applied to bodies undergoing large displacements and rotations require a large measurement area for both the reference (i.e., undeformed) image and the deformed images. This can limit the resulting resolution of the displacement and strain fields. To address this issue, we propose a two-stage dynamic DIC method capable of measuring displacements and strains under material convection with high resolution. During the first stage, the reference image is assembled from smaller, high-resolution images of the undeformed object obtained using a spatially-fixed or moving frame. Following capture, each sub-image is rigidly translated and rotated into its appropriate place, thereby producing a full, high-resolution image of the reference body. In stage two, images of the loaded and deformed body, again obtained using a small camera frame with high resolution, are aligned with matching regions of the undeformed composite image using BRISK feature detection before performing DIC. We demonstrate the method on a contact problem whereby an elastomeric roller travels along a rigid surface. In doing so, we obtain fine-resolution measurements of the state of strain of the region of the roller sidewall in contact with the substrate, even as new material convects through the region of interest. We present these measurements as a series of images and videos capturing strain evolution as the roller transitions from static loads to a fully dynamic steady-state, documenting the effectiveness of the method.