Search

Rail surface inspection using visual inspection system is an important part of railway maintenance. However, accurate and efficient identification of possible defects remains challenging. This paper proposes a background-oriented defect inspector (BODI) to improve defect detection by considering specified characteristics of the track during inspection. Reformulating the inspection task in this manner offers a new way to model rail surface images. More specifically, BODI features a random sampling stage to obtain a compact background representation without any prior information. A sufficient number of random selections generates adequate and diverse background statistics, and defect-determination and a fusion of procedures then determine whether current pixel belongs to the background. Finally, a background update mechanism and parallelism ensure real-time applicability. The proposed BODI is evaluated on a working railway line. The experimental results demonstrate that it outperforms state-of-the-art methods.

Computer vision systems have attracted much attention in recent years for use in detecting surface defects on rails; however, accurate and efficient recognition of possible defects remains challenging due to the variations shown by defects and also noise. This paper proposes a coarse-to-fine model (CTFM) to identify defects at different scales. The model works on three scales from coarse to fine: subimage level, region level, and pixel level. At the subimage level, the background subtraction model exploits row consistency in the longitudinal direction, and strongly filters the defect-free range, leaving roughly identified subimages within which defects may exist. At the next level, the region extraction model, inspired by visual saliency models, locates definite defect regions using phase-only Fourier transforms. At the finest level, the pixel subtraction model uses pixel consistency to refine the shape of each defect. The proposed method is evaluated using Type-I and Type-II rail surface defect detection data sets and an actual rail line. The experimental results show that CTFM outperforms state-of-the-art methods according to both the pixel-level index and the defect-level index.

Rail inspection based on visual inspection system (VIS) has drawn much attention recently, since VIS is automatic, fast, nondestructive, and objective. However, visual rail inspection is still a challenge to accurately identify possible defects involving a large variety of visual appearances. This paper presents an automatic visual inspection system to address these difficulties. Rail surface images are first obtained by the system, and then fed into the proposed hierarchical inspection framework containing coarse extractor and fine extractor. Specifically, the coarse extractor handles this defect inspection problem by exploring characteristics of rail background instead of defect itself, and focuses on finding the background modes. The fine extractor, which integrates the longitudinal context information and transversal prior information, is presented to largely reduce the impacts of other irregulars. The proposed method is evaluated on labeled Type-I and Type-II rail surface discrete defects data sets and a practical rail line. The experimental results demonstrate that our method outperforms the state-of-the-art works.

All staffs are strictly requested to wear hard-hats when working in substations. Various object detection algorithms, especially those based on deep learning, thus have been proposed for the corresponding purpose. A deep learning based-object detection algorithm commonly involves a fundamental neural network which dominates detection performances, so this paper investigates different types of networks’ applicability when utilizing them with a typical object detection algorithm for the monitoring of hard-hat wearing in substations. This is conducted from various perspectives concerned by real-world implementation that includes time consumption, computation speed, precision and more. As a consequence, this study provides a guideline to the selection of the most appropriate deep neural network architectures for the specific monitoring scenario.

Cloud type classification plays an essential role in ground-based cloud observation. However, it is a challenge to accurately identify the cloud categories that involve a large variety of structural patterns and visual appearances. Image representation and classifier are the crucial factors for cloud classification, though they are individually investigated in the literature. This paper proposes a new cloud type classification method using duplex norm-bounded sparse coding (DNSC), which designs image representation and classifier under the same framework, i.e., norm-bounded sparse coding (NSC). NSC is not only used to encode local descriptors, but also well explored to develop an effective classifier. NSC takes both locality and sparseness into account, and it can be benefit to capture discriminative patterns for image representation and have discriminative power for classifier. Furthermore, NSC has closed-form solution and can be computed efficiently. More specifically, DNSC first extracts local descriptor from an input cloud image, and then DNSC forms a holistic representation leveraging NSC and max-pooling strategy. Finally, a classifier is built on the holistic representation using NSC. The proposed DNSC is evaluated on the total-sky cloud image set, and the experimental results demonstrate that DNSC outperforms the state-of-the-art methods and its accuracy increases by about $7\%$ compared with baselines. In addition, the categorywise performance improvement is particularly pronounced over the complex categories, such as Cirriform and Mixed cloudiness.

Effective scene representation is a fundamental part of high-resolution scene classification systems. In this letter, we present a holistic scene representation method, i.e., the two-level feature representation (TLFR) model. The TLFR is composed of low-level and high-level features. Low-level features are obtained by computing the residual error between a local descriptor and its corresponding visual word, and the high-level features are obtained using a proposed selection-constrained sparse coding method. In addition, low-level features in a cluster are integrated by summation pooling, whereas high-level features are fused by maximization pooling. The holistic scene representation is finally generated by incorporating these two levels of features into the bag-of-visual-words framework. Experimental results show that the TLFR model is robust to translation and rotation variations and demonstrates promising performance with the Land Use and Land Cover Database data set and a newly released Singapore data set.

The lipid bilayer membrane is a major component of cells and intracellular membranes in vivo, and plays an indispensable role in organisms. Currently in the medical field, lipid bilayer membranes are used as devices for targeted delivery of proteins, nucleic acids and drugs in the treatment of a variety of diseases. Based on coarse-grained molecular dynamics (CGMD), this paper studies the modeling method of the lipid bilayer with water as solvent in the Martini force field. The coarse-grained lipid bilayer membrane model can expand the time scale and improves the computational efficiency of molecular dynamics (MD) simulation. The self-assembly process of DPPC bilayer lipid membrane is simulated in Martini force field and the radial distribution function (RDF) is also analyzed. In addition, this paper studies the molecular dynamics of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) lipid bilayer membrane electroporation under Martini force field, which provides a theoretical reference for the application of coarse-grained molecular dynamics in biomedical and other related fields.

[Image: see text] Covalent probes can display unmatched potency, selectivity, and duration of action; however, their discovery is challenging. In principle, fragments that can irreversibly bind their target can overcome the low affinity that limits reversible fragment screening, but such electrophilic fragments were considered nonselective and were rarely screened. We hypothesized that mild electrophiles might overcome the selectivity challenge and constructed a library of 993 mildly electrophilic fragments. We characterized this library by a new high-throughput thiol-reactivity assay and screened them against 10 cysteine-containing proteins. Highly reactive and promiscuous fragments were rare and could be easily eliminated. In contrast, we found hits for most targets. Combining our approach with high-throughput crystallography allowed rapid progression to potent and selective probes for two enzymes, the deubiquitinase OTUB2 and the pyrophosphatase NUDT7. No inhibitors were previously known for either. This study highlights the potential of electrophile-fragment screening as a practical and efficient tool for covalent-ligand discovery.

Covalent probes can display unmatched potency, selectivity and duration of action, however, their discovery is challenging. In principle, fragments that can irreversibly bind their target can overcome the low affinity that limits reversible fragment screening. Such electrophilic fragments were considered non-selective and were rarely screened. We hypothesized that mild electrophiles might overcome the selectivity challenge, and constructed a library of 993 mildly electrophilic fragments. We characterized this library by a new high-throughput thiol-reactivity assay and screened them against ten cysteine-containing proteins. Highly reactive and promiscuous fragments were rare and could be easily eliminated. By contrast, we found selective hits for most targets. Combination with high-throughput crystallography allowed rapid progression to potent and selective probes for two enzymes, the deubiquitinase OTUB2, and the pyrophosphatase NUDT7. No inhibitors were previously known for either. This study highlights the potential of electrophile fragment screening as a practical and efficient tool for covalent ligand discovery.

We report a novel platform using optimized SiO 2 @Au core–shell structures as matrices for highly efficient laser desorption/ionization mass spectrometry analysis of small biomolecules (MW 2 @Au nanoshells can achieve low detection-of-limits (~pmol–fmol) in mass spectrometry and selective laser desorption/ionization in bio-mixtures towards diverse small molecules. By further surface modification with aptamers, Apt-SiO 2 @Au nanoshells allowed simultaneously targeted enrichment and detection of kanamycin with a detection limit at 200pM. Our work not only starts new applications of SiO 2 @Au nanoshells in mass spectrometry, but also contributes to advanced analysis of either a group of small molecules or one target small molecule from complex bio-samples in a pre-designed manner for bio-diagnostics. From the Clinical Editor Existing methods for the detection of small molecules are often not sensitive enough. Here, the authors developed aptamer functionalized SiO2@Au nanoshells for use in mass spectrometry, with very low detection limits. The new platform appeared to be simple and efficient and should be applicable in detection of clinical samples.

It is known that the ability of humans to perform basic-level visual classification (e.g. birds versus dogs) develops well before their ability to perform fine-grained visual categorization (e.g. distinguish different sub-categories from Hooded Oriole to Scott Oriole). Similarly, computer vision research follows such trajectory. Basic-level image classification has made great progress with the help of deep learning models in recent years, while fine-grained image categorization is still facing a number of challenges and attracting more and more attention. In this paper, we review the recent progress in fine-grained image categorization. Starting from its definition, we give a brief introduction to some recent developments in fine-grained image categorization. After that, we elaborate different algorithms from strongly supervised learning and weakly supervised learning, and compare their performances on four publicly available benchmarks. Finally, we provide a brief summary of these methods as well as the potential future research directions, i.e. suggest to explore deeper neural networks and generative adversarial networks.

In this paper, we propose an effective approach to detect defects in fabrics. Based on the observation that fabric textures usually form a low-rank structure and the structure can be violated by the presence of defects, we formulate the task as a low-rank and sparse matrix decomposition problem. Moreover, the prior that defects tend to be continuous regions is considered in our model and the estimation of defect levels is properly solved by introducing an integration mechanism. Experimental results demonstrate that our proposed method can not only detect defects accurately but also have greater ability to preserve defect details than traditional approaches.

Determining the properties of proteins prior to purification saves time and labor. Here, we demonstrate a native mass spectrometry approach for rapid characterization of overexpressed proteins directly in crude cell lysates. The method provides immediate information on the identity, solubility, oligomeric state, overall structure, and stability, as well as ligand binding, without the need for purification.

Amino-functionalized macroporous silica foam (NH2 -MOSF) has been developed as a host reactor to realize highly efficient proteolysis in acidic solutions where normal tryptic reactions cannot occur. The digestion protocol consists simply of adding the functionalized NH2 -MOSF into the protein and trypsin solutions without altering the bulk pH or preloading the enzymes on the materials. With this protocol, digestion of sample fractions from LC can be efficiently realized in the acidic solutions directly. Digestion of a protein fraction extracted from rat liver tissue after LC separation was performed to illustrate this principle, where 103 proteins were successfully identified at pH 3 after 1.5 h of tryptic digestion.

A mass spectrometry signal amplification method is developed for the ultrasensitive and selective detection of low-abundance protein biomarkers by utilizing tag molecules on gold nanoparticles (AuNPs). EpCAM and thrombin as model targets are captured by specific aptamers immobilized on the AuNPs. With laser desorption/ionization time-of-flight mass spectrometry (LDI-TOF MS), the mass tag molecules are detected to represent the protein biomarkers. Benefiting from the MS signal amplification, the assay can achieve a limit of detection of 100 aM. The method is further applied to detect thrombin in fetal bovine serum and EpCAM in cell lysates to demonstrate its selectivity and feasibility in complex biological samples. With the high sensitivity and specificity, the protocol shows great promise for providing a new route to single-cell analysis and early disease diagnosis.

We for the first time demonstrate multi-functional magnetic particles based rare cell isolation combined with the downstream laser desorption/ionization mass spectrometry (LDI-MS) to measure the metabolism of enriched circulating tumor cells (CTCs). The characterization of CTCs metabolism plays a significant role in understanding the tumor microenvironment, through exploring the diverse cellular process. However, characterizing cell metabolism is still challenging due to the low detection sensitivity, high sample complexity, and tedious preparation procedures, particularly for rare cells analysis in clinical study. Here we conjugate ferric oxide magnetic particles with anti-EpCAM on the surface for specific, efficient enrichment of CTCs from PBS and whole blood with cells concentration of 6-100 cells per mL. Moreover, these hydrophilic particles as matrix enable sensitive and selective LDI-MS detection of small metabolites (MW

We present a strategy for imaging of elements in biological tissues using laser ablation (LA) mass spectrometry (MS), which was compared to laser ablation inductively coupled plasma (LA-ICP) MS. Both methods were adopted for quantitative imaging of elements in mouse kidney, as well as traumatic brain injury model tissue sections. MS imaging (MSI) employing LA provides quantitative data by comparing signal abundances of sodium from tissues to those obtained by imaging quantitation calibration standards of the target element applied to adjacent control tissue sections. LA-ICP MSI provided quantitative data for several essential elements in both brain and kidney tissue sections using a dried-droplet approach. Both methods were used to image a rat model of traumatic brain injury, revealing accumulations of sodium and calcium in the impact area and its peripheral regions. LA MSI is shown to be a viable option for quantitative imaging of specific elements in biological tissue sections.

A versatile protocol has been developed for large-scale characterization of hydrophobic membrane proteins based on the periodic mesoporous organosilica (PMO) acting as both an extractor for hydrophobic substrate capture and a nanoreactor for efficient in situ digestion. With introduction of organic groups in the pore frameworks and the presence of hydrophilic silanol groups on the surface, PMO can be well-dispersed into not only an organic solution to concentrate the dissolved membrane proteins but also an aqueous solution containing enzymes for sequential rapid proteolysis in the nanopores. The unique amphiphilic property of PMO ensures a facile switch in different solutions to realize the processes of substrate dissolution, enrichment, and digestion effectively. Furthermore, this novel PMO-assisted protocol has been successfully applied for identification of complex membrane proteins extracted from mouse liver as proof of general applicability.

An electrochemical approach is described to assess the activities of thrombin and one of its inhibitors. Thrombin is a vital proteinase, which is one kind of serine protease, and plays an important role in coagulation cascade, thrombosis, and hemostasis. A gold electrode immobilized with p-aminodiphenylamine (pADA) modified peptide (H-d-Pro-Phe-Arg-p-aminodiphenylamine) was prepared for the analyses of thrombin and its inhibitor. The peptide was employed as the recognition and cleavage site for thrombin while pADA was used as an electroactive reporter. Experimental results showed that this method is sensitive as a low concentration of 5 fM thrombin can be detected. The inhibition of thrombin activity by argatroban was monitored and IC50 was found to be about 10 μM. The protocol was further applied for the measurement of a healthy human serum sample. Keywords: Thrombin, Argatroban, Inhibition, Electrochemistry, Sensor

A novel multifunctional nanoreactor was designed for highly efficient detection and profiling of membrane proteins using macroporous siliceous foams (MSF) with large pore size of ∼ 70 nm and hydrophobic surface (contact angle over 100°). By introducing organo groups (e. g. from bis(triethoxysilyl)ethane) on the surface of MSF, the modified material was employed for selective extraction and in situ tryptic digestion of membrane proteins coupled with mass spectrometry (MS) based profiling in biological samples. The optimized nanoreactor was selected in detection of a standard membrane protein (bacteriorhodopsin) and further applied to Hela cell lysates yielding 112 membrane proteins identified. Our nanoreactors shed light on MS analysis of membrane proteins towards biomarkers discovery and clinical diagnostics.

We report a novel platform using optimized SiO2@Au core-shell structures as matrices for highly efficient laser desorption/ionization mass spectrometry analysis of small biomolecules (MW700 Da). Owing to the designer structure, SiO2@Au nanoshells can achieve low detection-of-limits (~pmol-fmol) in mass spectrometry and selective laser desorption/ionization in bio-mixtures towards diverse small molecules. By further surface modification with aptamers, Apt-SiO2@Au nanoshells allowed simultaneously targeted enrichment and detection of kanamycin with a detection limit at 200 pM. Our work not only starts new applications of SiO2@Au nanoshells in mass spectrometry, but also contributes to advanced analysis of either a group of small molecules or one target small molecule from complex bio-samples in a pre-designed manner for bio-diagnostics.Existing methods for the detection of small molecules are often not sensitive enough. Here, the authors developed aptamer functionalized SiO2@Au nanoshells for use in mass spectrometry, with very low detection limits. The new platform appeared to be simple and efficient and should be applicable in detection of clinical samples.

An advanced approach is developed in this work for simultaneous on-line separation and digestion of proteins by combining the Off-Gel isoelectric focusing (IEF) and enzymatic nanoreactor enhanced proteolysis. The nanoreactor was prepared by preloading trypsin in amino-functionalized macroporous silica, and then directly added into Off-Gel wells. With the nanoreactor loaded Off-Gel device, effective digestion of proteins happened during IEF electrophoresis to generate directly fractionated tryptic peptides, which not only accelerated the experimental flow but also avoided sample loss, leading to a more comprehensive protein identification from complex biological samples. A successful identification of 3592 proteins was achieved from Hela cell line by using the approach followed with LC-MS/MS analysis, while only 1877 proteins were identified from the same sample when using standard in-solution proteolysis followed with Off-Gel electrophoresis and then LC-MS/MS analysis. Therefore, we have demonstrated that the approach can greatly simplify high-throughput proteomics and significantly improve protein identification. (C) 2015 Elsevier B.V. All rights reserved.