YOLO-DD:Improved YOLOv5 for Defect Detection

As computer technology continues to advance,factories have increasingly higher demands for detecting defects.However,detecting defects in a plant environment remains a challenging task due to the presence of complex backgrounds and defects of varying shapes and sizes.To address this issue,this paper proposes YOLO-DD,a defect detectionmodel based on YOLOv5 that is effective and robust.To improve the feature extraction process and better capture global information,the vanilla YOLOv5 is augmented with a new module called Relative-Distance-Aware Transformer(RDAT).Additionally,an Information Gap Filling Strategy(IGFS)is proposed to improve the fusion of features at different scales.The classic lightweight attention mechanism Squeeze-and-Excitation(SE)module is also incorporated into the neck section to enhance feature expression and improve the model’s performance.Experimental results on the NEU-DET dataset demonstrate that YOLO-DDachieves competitive results compared to state-of-the-art methods,with a 2.0% increase in accuracy compared to the original YOLOv5,achieving 82.41% accuracy and38.25FPS(framesper second).Themodel is also testedon a self-constructed fabric defect dataset,and the results show that YOLO-DD is more stable and has higher accuracy than the original YOLOv5,demonstrating its stability and generalization ability.The high efficiency of YOLO-DD enables it to meet the requirements of industrial high accuracy and real-time detection.

Analytical Modeling and Mechanism Analysis of Time-Varying Excitation for Surface Defects in Rolling Element Bearings

Surface defects,including dents,spalls,and cracks,for rolling element bearings are the most common faults in rotating machinery.The accurate model for the time-varying excitation is the basis for the vibration mechanism analysis and fault feature extraction.However,in conventional investigations,this issue is not well and fully addressed from the perspective of theoretical analysis and physical derivation.In this study,an improved analytical model for time-varying displacement excitations(TVDEs)caused by surface defects is theoretically formulated.First and foremost,the physical mechanism for the effect of defect sizes on the physical process of rolling element-defect interaction is revealed.According to the physical interaction mechanism between the rolling element and different types of defects,the relationship between time-varying displacement pulse and defect sizes is further analytically derived.With the obtained time-varying displacement pulse,the dynamic model for the deep groove bearings considering the internal excitation caused by the surface defect is established.The nonlinear vibration responses and fault features induced by surface defects are analyzed using the proposed TVDE model.The results suggest that the presence of surface defects may result in the occurrence of the dual-impulse phenomenon,which can serve as indexes for surface-defect fault diagnosis.

Attention-relation network for mobile phone screen defect classification via a few samples

How to use a few defect samples to complete the defect classification is a key challenge in the production of mobile phone screens.An attention-relation network for the mobile phone screen defect classification is proposed in this paper.The architecture of the attention-relation network contains two modules:a feature extract module and a feature metric module.Different from other few-shot models,an attention mechanism is applied to metric learning in our model to measure the distance between features,so as to pay attention to the correlation between features and suppress unwanted information.Besides,we combine dilated convolution and skip connection to extract more feature information for follow-up processing.We validate attention-relation network on the mobile phone screen defect dataset.The experimental results show that the classification accuracy of the attentionrelation network is 0.9486 under the 5-way 1-shot training strategy and 0.9039 under the 5-way 5-shot setting.It achieves the excellent effect of classification for mobile phone screen defects and outperforms with dominant advantages.

Unraveling abnormal buried interface anion defect passivation mechanisms depending on cation-induced steric hindrance for efficient and stable perovskite solar cells

Although ionic liquids(ILs)have been widely employed to heal the defects in perovskite solar cells(PSCs),the corresponding defect passivation mechanisms are not thoroughly understood up to now.Herein,we first reveal an abnormal buried interface anion defect passivation mechanism depending on cationinduced steric hindrance.The IL molecules containing the same anion([BF4]^(-))and different sizes of imidazolium cations induced by substituent size are used to manipulate buried interface.It was revealed what passivated interfacial defects is mainly anions instead of cations.Theoretical and experimental results demonstrate that the large-sized cations can weaken the ionic bond strength between anions and cations,and facilitate the interaction between anions and SnO2as well as perovskites,which is conducive to interfacial defect passivation and ameliorating interfacial contact.It can be concluded that interfacial chemical interaction strength and defect passivation effect are positively correlated with the size of cations.The discovery breaks conventional thinking that large-sized modification molecules would weaken their chemical interaction with perovskite.Compared with the control device(21.54%),the device based on 1,3-Bis(1-adamantyl)-imidazolium tetrafluoroborate(BAIMBF4)with maximum size cations achieves a significantly enhanced efficiency of 23.61%along with much increased moisture,thermal and light stabilities.

Perfecting HER catalysts via defects:Recent advances and perspectives

Defect engineering has become a promising approach to improve the performance of hydrogen evolution reaction(HER)catalysts.Non-noble transition metal-based catalysts(TMCs)have shown significant promise as effective alternatives to traditional platinum-group catalysts,attracting considerable attention.However,the industrial application of TMCs in electrocatalytic hydrogen production necessitates further optimization to boost both catalytic activity and stability.This review comprehensively examines the types,fabrication methods,and characterization techniques of various defects that enhance catalytic HER activity.Key advancements include optimizing defect concentration and distribution,coupling heteroatoms with vacancies,and leveraging the synergy between bond lengths and defects.In-depth discussions highlight the electronic structure and catalytic mechanisms elucidated through in-situ characterization and density functional theory calculations.Additionally,future directions are identified,exploring novel defect types,emphasizing precision synthesis methods,industrial-scale preparation techniques,and strategies to enhance structural stability and understanding the in-depth catalytic mechanism.This review aims to inspire further research and development in defect-engineered HER catalysts,providing pathways for high efficiency and cost-effectiveness in hydrogen production.

A Simple and Effective Surface Defect Detection Method of Power Line Insulators for Difficult Small Objects

Insulator defect detection plays a vital role in maintaining the secure operation of power systems.To address the issues of the difficulty of detecting small objects and missing objects due to the small scale,variable scale,and fuzzy edge morphology of insulator defects,we construct an insulator dataset with 1600 samples containing flashovers and breakages.Then a simple and effective surface defect detection method of power line insulators for difficult small objects is proposed.Firstly,a high-resolution featuremap is introduced and a small object prediction layer is added so that the model can detect tiny objects.Secondly,a simplified adaptive spatial feature fusion(SASFF)module is introduced to perform cross-scale spatial fusion to improve adaptability to variable multi-scale features.Finally,we propose an enhanced deformable attention mechanism(EDAM)module.By integrating a gating activation function,the model is further inspired to learn a small number of critical sampling points near reference points.And the module can improve the perception of object morphology.The experimental results indicate that concerning the dataset of flashover and breakage defects,this method improves the performance of YOLOv5,YOLOv7,and YOLOv8.In practical application,it can simply and effectively improve the precision of power line insulator defect detection and reduce missing detection for difficult small objects.

PdCu alloy anchored defective titania for photocatalytic conversion of carbon dioxide into methane with 100% selectivity

The photoreduction of CO_(2)into CH_(4)with simultaneous high activity and selectivity is a promising strategy to increase energy supply and alleviate global warming.However,the absence of the active sites that is responsible for the adsorption and activation of CO_(2)and the generation of CO and H2via side reactions often lead to poor efficiency and low selectivity of the catalyst.Herein,Cu,Pd,and PdCu metal clusters cocatalyst-anchored defective TiO_(2)nanotubes(Cu/TiO_(2)-SBO,Pd/TiO_(2)-SBO,and Pd1Cu1/TiO_(2)-SBO)were designed via a simple solution impregnation reduction and applied for photocatalytic conversion of CO_(2)to CH_(4).The Pd1Cu1/TiO_(2)-SBO photocatalyst exhibits excellent catalytic performance among the other catalysts for photoreduction of CO_(2)into CH_(4).More interestingly,the product selectivity of CH_(4)reaches up to 100%with a rate of 25μmol g^(-1)h^(-1).In-situ diffuse reflectance infrared Fourier transform spectroscopy(DRIFTS)and density functional theory(DFT)simulations indicate that the main reasons for the high selectivity of CH_(4)are attributed to the PdCu alloy and oxygen vacancies,which jointly enhance the photoinduced carrier separation and lower energy barriers of key intermediates.Moreover,due to the tunable d-band center of the Cu site in the PdCu alloy,the generated intermediates can be well prevented from poisoning and promoted to participate in further reactions.Hopefully,the current study will provide insight into the development of new,highly selective photocatalysts for the visible light-catalytic reduction of CO_(2)into CH_(4).

A review of β-Ga_2O_3 single crystal defects, their effects on device performance and their formation mechanism

As a wide-bandgap semiconductor(WBG), β-Ga_2O_3 is expected to be applied to power electronics and solar blind UV photodetectors. In this review, defects in β-Ga_2O_3 single crystals were summarized, including dislocations, voids, twin, and small defects. Their effects on device performance were discussed. Dislocations and their surrounding regions can act as paths for the leakage current of SBD in single crystals. However, not all voids lead to leakage current. There's no strong evidence yet to show small defects affect the electrical properties. Doping impurity was definitely irrelated to the leakage current. Finally, the formation mechanism of the defects was analyzed. Most small defects were induced by mechanical damages. The screw dislocation originated from a subgrain boundary. The edge dislocation lying on a plane slightly tilted towards the(102) plane, the(101) being the possible slip plane. The voids defects like hollow nanopipes, PNPs, NSGs and line-shaped grooves may be caused by the condensation of excess oxygen vacancies, penetration of tiny bubbles or local meltback. The nucleation of twin lamellae occurred at the initial stage of "shoulder part" during the crystal growth. These results are helpful in controlling the occurrence of crystal defects and improving the device performance.

DEBINDING MECHANISM AND KINETICS FOR PW IN PW－WC FEEDSTOCK UNDER AIR ATMOSPHERE

ＤＥＢＩＮＤＩＮＧＭＥＣＨＡＮＩＳＭＡＮＤＫＩＮＥＴＩＣＳＦＯＲＰＷＩＮＰＷ－ＷＣＦＥＥＤＳＴＯＣＫＵＮＤＥＲＡＩＲＡＴＭＯＳＰＨＥＲＥＬｉａｎｇＳｈｕｑｕａｎ；ＨｕａｎｇＢａｉｙｕｎ；ＱｕＸｕａｎｈｕｉ（ＮａｔｉｏｎａｌＬａｂｏｒａｔｏｒｙｆｏｒＰ...

Defect Formation Mechanisms in Selective Laser Melting:A Review

Defect formation is a common problem in selective laser melting （SLM）. This paper provides a review of defect formation mechanisms in SLM. It sum- marizes the recent research outcomes on defect findings and classification, analyzes formation mechanisms of the common defects, such as porosities, incomplete fusion holes, and cracks. The paper discusses the effect of the process parameters on defect formation and the impact of defect formation on the mechanical properties of a fabri- cated part. Based on the discussion, the paper proposes strategies for defect suppression and control in SLM.

Forming Mechanism of Gaseous Defect in Ti-48Al-2Cr-2Nb Exhaust Valves Formed with Permanent Mold Centrifugal Casting Method

A method combining theoretical analysis with experiment is adopted and the flowing process of Ti-48A1-2Cr-2Nb alloy melt poured in a permanent mould during the centrifugal casting process has been analyzed. A mathematical model of the filling process is established and the forming mechanism of internal gaseous defect is summarized. The results of calculation show that the melt fills the mould with varying cross-section area and inclined angle. The filling speed of the cross-section is a function of filling time. The cross-section area is directly proportional to the filling speed and the inclined angle is inversely proportional to the filling speed at a given rotating speed of the platform. Both of them changes more obvious near the mould entrance. The gaseous defect can be formed in several ways and the centrifugal field has an important influence on the formation of the defect. In addition, the filling process in centrifugal field has been verified by wax experiments and the theoretical analysis are consistent with experimental results.

The Formation Mechanism and Control Methods of High Temperature Infrared Radiant Coating Defects

Using high aluminum refractory material as substrate at 1400℃, we studied the connections between several oxides such as Fe203, MnOv CuO, and the formation of defects such as coating crack, exfoliation, blistering, erosion, and fading away appeared in the application of high temperature infrared radiation coating. Analyses showed that thermal stress formed during the heating process due to the thermal expansion coefficient differential between the coating and the substrate, and volume effect caused by the crystal transferred when the temperature changed, which resulted in the coating crack and exfoliation. The gas produced by the reactions between components and binder or the components themselves during the heating process caused the coating blistering. The EMPA and XRD analyses show that oxides with low melting point in the penetrating area of the substrate may form eutectic with low melting point and produced thermal defects, which leads to the erosion by penetrating to the substrate. The valent changes of Fe2O3 and MnO2 during the heating process cause the volatilization of the oxides or the pulverization of the coatings, resulting in the coating fades away easily at high temperature for a long time.

Formation and Removement Mechanism of Haze Defects on（111）p－type Silicon Wafers

The haze defects on p-type (111) silicon wafers were investigated by means of chemical etching, Fouriertransform infra-red microscopy (FTIR), spreading resistance measurement. secondary ion mass spectroscopy(SLMS), transmission electron microscopy (TEM) equipped with an energy-dispersive X-ray spectrometer(EDX). The haze defects are the precipitates of silicide of metal impurities (Fe, Ni) on the wafer surface.The formation of haze defects can efficiently be inhibited by utilizing the technology of fast neutronirradiation combined with the internal gettering (IG), and then, the formation and removement mechanismof the haze defects have been discussed in this paper.

Surface defects,stress evolution,and laser damage enhancement mechanism of fused silica under oxygen-enriched condition

Oxygen ions(O;)were implanted into fused silica at a fixed fluence of 1×10^(17) ions/cm^(2) with different ion energies ranging from 10 ke V to 60 ke V.The surface roughness,optical properties,mechanical properties and laser damage performance of fused silica were investigated to understand the effect of oxygen ion implantation on laser damage resistance of fused silica.The ion implantation accompanied with sputtering effect can passivate the sub-/surface defects to reduce the surface roughness and improve the surface quality slightly.The implanted oxygen ions can combine with the structural defects(ODCs and E′centers)to reduce the defect densities and compensate the loss of oxygen in fused silica surface under laser irradiation.Furthermore,oxygen ion implantation can reduce the Si-O-Si bond angle and densify the surface structure,thus introducing compressive stress in the surface to strengthen the surface of fused silica.Therefore,the laser induced damage threshold of fused silica increases and the damage growth coefficient decreases when ion energy up to30 ke V.However,at higher ion energy,the sputtering effect is weakened and implantation becomes dominant,which leads to the surface roughness increase slightly.In addition,excessive energy aggravates the breaking of Si-O bonds.At the same time,the density of structural defects increases and the compressive stress decreases.These will degrade the laser laser-damage resistance of fused silica.The results indicate that oxygen ion implantation with appropriate ion energy is helpful to improve the damage resistance capability of fused silica components.

Healing Mechanism and Osteogenic Capacity of Bovine Bone Mineral—Human Amniotic Mesenchymal Stem Celland Autogenous Bone Graft in Critical Size Mandibular Defect

Experiments on maxillofacial bone tissue engineering showed the promising result;however, its healing mechanisms and effectiveness had not been fully understood. The aim of this study is to compare the bone healing mechanism and osteogenic capacity between bovine bone mineral loaded with hAMSC and autogenous bone graft in the reconstruction of critical size mandibular bone defect. Critical size defects were made at the mandible of 45 New Zealand white rabbits reconstructed with BBM-hAMSC, BBM alone, and ABG, respectively. At the end of first, second, and twelfth weeks, five rabbits from each experimental week were sacrificed for histology and immunohistochemistry staining. Expressions of vascular endothelial growth factor (VEGF), bone mor-phogenic proteins-2 (BMP2), Runx2 and the amount of angiogenesis were analyzed in the first and second week groups, while expressions of Runx2, osteocalcin, collagen type-I fibres, trabecular area and bone incorporation were analyzed in the twelfth week groups. The result showed that expressions of VEGF, BMP2 and Runx2 as well as the amount of angiogenesis were higher in ABG compared with BBM-hAMSC group in the first and second weeks of healing. The result of twelfth week of healing showed that expressions of Runx2 and osteocalcin as well as the thickness of collagen type-I fibres were significantly higher in BBM-hAMSC compared to ABG group, while there was no statistically difference in trabecular area and bone incorporation between BBM-hAMSC and ABG group. This study concluded that early healing activities were higher in auto-genous bone graft than in BBM-hAMSC, while osteogenic activities in the late stage of healing were higher in BBM-hAMSC compared to autogenous bone graft. It was also concluded that the osteo-genic capacity of BBM-hAMSC was comparable to autogenous bone graft in the reconstruction of critical size defect in the mandible.

Possible mechanism of 15D-PGJ2 in promoting periodontal tissue regeneration in patients with mandibular defects

Objective:To explore the main physiological mechanism of 15d-PGJ2 promoting periodontal tissue regeneration in patients with jaw defects caused by periodontal disease.Methods:From February 2016 to July 2019,a controlled study was conducted on 73 healthy residents(healthy group)and 73 patients(case group)with periodontal disease combined with jaw defects in Changsha medical university.T test was used to compare the growth factors of gingival crevicular fluid between the two groups.Peripheral blood cells;Cement-specific protein;Peripheral blood enzyme;Statistical differences in bone metabolites.The t test method compared the content of each index before and after treatment(15d-PGJ2 was treated at a dose of 200 mu/kg for 14 days).The method of factor analysis explores the internal correlation of each index.Result:RANKL,ICAM-1,TGF-β1,Th17,Treg,PDLSCs,SOST,CAP,HMGB1,CTSK,5-LOX,COX-2,NTX were higher in the case group than in the healthy group.In the case group,RANKL,ICAM-1,TGF-β1,Th17,Treg,PDLSCs,SOST,CAP,HMGB1,CTSK,5-LOX,COX-2,NTX were lower than those in the healthy group.The differences between the groups were statistically significant(P<0.05).Compared with before treatment,IL-1β,IL-17,Bfgf,YKL-40,BMP-2,ICTP,PICP,CTX were significantly decreased after treatment.RANKL,ICAM-1,TGF-β1,Th17,Treg,PDLSCs,SOST,CAP,HMGB1,CTSK,5-LOX,COX-2,NTX were significantly increased.The differences were statistically significant(P<0.05).Factor analysis shows that four common factors can be extracted from 21 indicators,and the cumulative contribution rate is 96.993%.Conclusions:The treatment of 15d-PGJ2 in patients with periodontal disease with maxillary defects can significantly affect the expression of multiple characteristic indicators,which may involve four mechanisms:dysregulation of cell differentiation or migration,local inflammation or immune imbalance,destruction of alveolar bone microstructure,load or stimulation,and remodeling.The specific pathway related to this is still to be further explored.

注意力可变形卷积网络的木质板材瑕疵识别

为了解决木材缺陷检测中人工成本高、效率低的问题,该文基于可变性卷积网络和注意力机制,提出一种端到端的神经架构模型.首先,可变形卷积网络(Deformable Convolutional Network, DCN)通过将矩形网格转换为变形网格,使模型专注于具有更多有用图像信息的区域.使用可变形卷积网络可以忽略图像特征中不相关的系数,解决了传统卷积在特征中学习更多信息能力有限的问题.然后,将DCN输出馈送到门控循环单元(Gated Recurrent Unit, GRU)层以学习缺陷图像的高级特征.最后,通过关注输入图像的最重要特征,应用注意力机制加强瑕疵区域的高亮度,从而提高模型识别的准确性.使用Matlab平台在4个木质板材缺陷数据集上将该文方法与现有其他方法进行比较分析,该文方法的准确率比其他3种对比方法提高了2.4%~13.2%的维度,灵敏度提高了3.3%~16.6%的维度,特异性提高了4%~21%的维度.实验结果表明,该文方法在检测精度和其他各个性能方面均优于现有方法,最佳准确率为99.2%,证明了该文方法的有效性.

基于多尺度卷积注意力机制的输电线路防振锤缺陷检测

作为输电线路中的重要金具部件,防振锤的缺陷将对输电线路构成严重威胁。针对由于防振锤缺陷样本数量稀少、背景复杂、区域形状尺寸不一造成的防振锤缺陷识别能力不足的问题,提出一种基于多尺度卷积注意力机制的防振锤缺陷检测方法。首先,通过统计不同缺陷的防振锤尺寸,设计适应不同类别的多尺度卷积注意力机制,使网络重点关注图像中的防振锤区域;其次,引入结构重参数化方法,以将网络中的多分支结构无损失地转换为单分支结构,在提高网络检测性能的同时维持检测速度在较高水平;最后,以渐进式特征金字塔网络结构(AFPN)为基础,融合更多的浅层网络,提高了网络检测防振锤小目标的能力。实际收集的防振锤缺陷数据集实验结果表明,设计的检测方法可显著提升防振锤缺陷检测的性能,检测精度mAP0.5达到了91.9%,在TITAN XP平台下检测速度达60.88帧/s,可为输电线路防振锤智能化巡检提供参考。

变频电流激励下GIS隔离开关机械缺陷的振动非线性行为及其辨识方法

气体绝缘组合电器(gas insulated switchgear,GIS)机械缺陷是导致设备故障的关键因素之一,为此开展了变频电流激励模式下隔离开关底座松动缺陷的时频域振动图谱变化规律研究,进行了波形因子、重心频率等6个振动特征量的量化分析,采用灰狼算法优化的支持向量机研究提出了变频电流激励模式下GIS隔离开关底座松动缺陷严重程度辨识方法。结果表明:GIS隔离开关缺陷振动信号特征量与电流激励频率具有强相关性,与单一工频电流激励模式相比,变频电流激励具备更有效检测GIS设备机械缺陷的优势;基于126 kV真型GIS试验平台验证获得构建的隔离开关底座松动缺陷辨识模型的平均识别准确率达90%以上。该研究成果为揭示GIS设备的非线性机械振动特性及提升其机械缺陷检测水平提供了重要参考。

IDD-YOLOv7:一种用于输电线路绝缘子多缺陷的轻量化检测方法

YOLO目标检测算法是当前基于图像的输电线路绝缘子缺陷检测的主流方法,然而现有模型复杂度较大,亟需合理有效的参数压缩方法作为前提条件,来为解决无人机边缘设备部署的困境问题奠定基础;同时,无人机航拍的绝缘子缺陷图像背景复杂、缺陷尺寸较小,容易出现误检、漏检等问题。为此,提出了一种用于输电线路绝缘子多缺陷检测的Insulator Defect Detection-YOLOv7(IDD-YOLOv7)模型,以降低模型复杂度,提高模型鲁棒性。首先,在多尺度特征融合的过程中加入坐标注意力(Coordinate Attention)机制,抑制复杂背景的干扰,提升模型对小目标的全局感知能力;之后,设计C3GhostNetV2模块,用于捕获不同空间像素之间的远程依赖性,在增强模型表达能力的同时降低模型的参数量和浮点运算量;最后,提出Focal-CIoU损失函数,提高模型高质量anchor的贡献,加快模型的收敛速度。实验结果表明,本文方法与基线模型相比mAP^(50)提升了3.8%,查准率和召回率分别提升了1.7%和7.6%,参数量和浮点运算量分别下降了18.3%和14.0%,绝缘子自爆、破损、闪络缺陷的AP^(50)分别提升了0.8%、4.5%、6.3%。