检索结果-内蒙古大学图书馆

Automatic welding imperfections detection in a smart factory via 2-D laser scanner

JOURNAL OF MANUFACTURING PROCESSES 2022年 73卷 948-960页

作者： Bologna, Francesco Tannous, Michael Romano, Donato Stefanini, Cesare Scuola Super Sant Anna BioRobot Inst Viale R Piaggio 34 I-56025 Pontedera PI Italy Scuola Super Sant Anna Dept Excellence Robot & Artificial Intelligence I-56025 Pisa Italy Baker Hughes Co Nuovo Pignone Tecnol Srl Via Felice Matteucci 2 I-50127 Florence FI Italy Khalifa Univ Healthcare Engn Innovat Ctr Abu Dhabi U Arab Emirates

Welding defect detection is crucial to determine welding quality and assure correct mechanical performance of the welded joint during operation. Most of common detection techniques are performed offline, are time consuming, and strongly dependent on the operator skill. In the present work, two novel automatic online welding imperfections detection approaches are presented for classifying qualified and defective profiles acquired with a 2-D laser scanner on square grooves in a multi-pass Gas Tungsten Arc Welding process. In particular, a Heuristic Algorithm has been developed to analyze different geometrical characteristics of the profile for its classification. The Heuristic Algorithm has been validated on five welded coupons, created in Baker Hughes welding laboratory in Massa, Italy. Coupons have been scanned and resulting profiles have been manually labeled. The Heuristic Algorithm showed good performance, with a F1 score of 94.8%. Additionally, a multi-layer feed forward Neural Network has been trained and tested on the profiles acquired from the five welded coupons, with an overall F1 score of 91.0%. The promising results of the two proposed approaches will lead to a wider test campaign, before implementing a mixed solution in a multi-pass robotic Gas Tungsten Arc Welding industrial process.

关键词： Artificial intelligence Automation detection algorithms Industry 4 0 Welding

来源：评论

学校读者我要写书评

暂无评论

GF-4 Satellite Fire detection With an Improved Contextual Algorithm

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2022年 15卷 163-172页

作者： Zhang, Ning Sun, Lin Sun, Zhendong Shandong Univ Sci & Technol Coll Geodesy & Geomat Qingdao 266590 Peoples R China Beijing Res Ctr Informat Technol Agr Key Lab Quantitat Remote Sensing Agr Minist Agr Beijing 100097 Peoples R China

The GF-4 satellite has a moderate spatial resolution and is capable of high temporal frequency observation. Thus, it can play an important role in the acquisition of information regarding fires. In this article, the characteristics of brightness temperature difference between fire point area and nonfire point area are comprehensively analyzed. Employing the advantages of high-frequency continuous observation of these data, an improved algorithm for joint spatial attribute detection using multitemporal data is proposed. This method is based on the traditional contextual method and incorporates the concept of time sequence, and uses the changes in time phase before and after the fire to reduce the false alarm caused by the high heterogeneity of pixels in a single spatial scene in the traditional method. In order to evaluate the accuracy of the algorithm, the perfect moderate resolution imaging spectrometer thermal anomaly product is used as the verification data to evaluate the feasibility of the algorithm. At the same time, it is necessary to compare the algorithm in this article with the fire algorithm currently applied to GF-4 with the four typical regions. The improved algorithm shows better drawing ability, and the consistency of fire description of biomass combustion for many days is improved. In the case of the Heilongjiang region on October 30, 2017, the accuracy rate is improved by 62%, compared with the contextual method.

关键词： Reflectivity Temperature measurement MODIS Combustion Brightness temperature Monitoring detection algorithms Brightness temperature fire point identification GF-4 spatial characteristics time phase change

来源：评论

学校读者我要写书评

暂无评论

YOLO8-FASG: A High-Accuracy Fish Identification Method for Underwater Robotic System

引用

IEEE ACCESS 2024年 12卷 73354-73362页

作者： Qin, Xiangrong Yu, Changdong Liu, Baisheng Zhang, Zhihao Dalian Maritime Univ Coll Informat Sci & Technol Dalian 116026 Peoples R China Dalian Maritime Univ Coll Artificial Intelligence Dalian 116026 Peoples R China Liaoning Tech Univ Coll Software Huludao 125000 Peoples R China

In underwater robots, accurate identification of small fish is still a major challenge, because small fish move faster and occupy less screen space, which requires higher detection flexibility and receptive field of the model. To solve this challenge, we propose a high-precision small fish identification and tracking method named YOLO8-FASG in this paper. Specifically, the proposed method is improved in three aspects based on the YOLOv8 framework. First, Alterable Kernel Convolution (AKConv) is used in the neck network of the model to automatically adjust the shape of the convolution kernel according to the size and shape of the object. In this way, the shape and contour characteristics of rapidly changing fish can be captured more accurately and efficiently;Second, we introduce a global attention mechanism (GAM) to broaden the receptive field of the model by enhancing attention to fish features from the two dimensions of channel and space;Third, we employ Simplified Spatial Pyramid Pooling-Fast (SimSPPF) to replace the standard Spatial Pyramid Pooling-Fast (SPPF) to enhance prediction accuracy. These improvements enable the model to effectively extract image features of small, fast fish, thereby improving the robot's accuracy in identifying small fish underwater. Experiments results in the public dataset Fish4Knowledge show that YOLO8-FASG performs significantly better than traditional YOLOv8 in underwater environments. Specifically, Precision and Recall increased by 1.6% and 3.5% respectively, while mAP50 and mAP50-95 increased by 1.3% and 6.1% respectively, and our method provides an effective solution for underwater robots to identify fish schools.

关键词： Feature extraction Task analysis Convolutional neural networks YOLO Object detection Predictive models Autonomous underwater vehicles Mobile robots Fish schools detection algorithms Spatial resolution YOLOv8 global attention mechanism AkConv SimSPPF object detection

来源：评论

学校读者我要写书评

暂无评论

A 40-nm 131-mW 6.4-Gb/s 256 x 32 Multi-User MIMO OTFS Detector for Next-Gen Communication Systems

引用

IEEE JOURNAL OF SOLID-STATE CIRCUITS 2025年

作者： Lee, Tang Chen, Ting-Yang Liu, I-Hsuan Yang, Chia-Hsiang Natl Taiwan Univ Grad Inst Elect Engn Taipei 10617 Taiwan Natl Taiwan Univ Dept Elect Engn 10617 10617 Taiwan

High-mobility communication technology enables important applications in the near future. In such scenarios, the wireless channel exhibits high Doppler spread, which makes orthogonal frequency division multiplexing (OFDM) adopted in current wireless system suffer from severe inter-carrier interference (ICI). The orthogonal time frequency space (OTFS) technique is a promising modulation to address this issue. It demonstrates higher resilience to Doppler spread than OFDM in terms of bit error rate (BER) at the expense of higher detection complexity. This work presents the first high-throughput multi-user multi-input multi-output (MU-MIMO) detector for OTFS communication systems. A low-complexity message-passing (MP) detection algorithm is proposed to achieve 93% lower computational complexity by leveraging the structure of Gram matrix. A memory-efficient residual noise (RN) update scheme is devised to reduce the memory size for storing the partial interference by 94%. The proposed MP detector achieves a 60% reduction in latency by employing mean computation unit (MCU) and dual-mode multiplier. In addition, a 91% memory access reduction and an 89% memory size reduction in the channel memory bank are achieved, respectively, by leveraging layer ordering, partial Gram matrix saving, and block diagonal approximation. The chip supports up to 32 users, 256 receive antennas, and 256-QAM modulation. Fabricated in a 40-nm CMOS technology, the chip integrates 6.76 M gates in area of 6.47 mm(2) and it delivers a maximal throughput of 6.4 Gb/s. The power consumption is 131 mW at 200 MHz from a 0.9 V supply. In comparison to state-of-the-art MU-MIMO OFDM detectors, this work achieves 3.3-to-21.3x higher maximal throughput and 2.2-to-67.0x lower normalized energy, in addition to higher resilience to Doppler spread.

关键词： Detectors OFDM Symbols Doppler effect Modulation Vectors Wireless communication Bit error rate Interference detection algorithms Algorithm-architecture co-optimization digital integrated circuit message-passing (MP) detector multi-user multi-input multi-output (MU-MIMO) system orthogonal time frequency space (OTFS) modulation

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Transfer Learning, LeafNet, and Modified LeafNet Models for Accurate Rice Leaf Diseases Classification

引用

IEEE ACCESS 2024年 12卷 36622-36635页

作者： Altabaji, Wassem I. A. E. Umair, Muhammad Tan, Wooi-Haw Foo, Yee-Loo Ooi, Chee-Pun Multimedia Univ Fac Engn Cyberjaya 63100 Selangor Malaysia

Early detection of plant diseases is essential for effective crop disease management to prevent yield loss. In this study, we developed a methodology for classifying diseases in rice leaves using four deep learning models and a dataset with 2658 images of healthy and diseased rice leaves. Four models, namely LeafNet, Modified LeafNet, MobileNetV2, and Xception, were compared. The Modified LeafNet model involved updates to LeafNet's architectural parameters, whereas transfer learning techniques were applied to the MobileNetV2 and Xception pretrained models. The optimal hyperparameters for training were determined by considering several factors such as batch size, data augmentation, learning rate, and optimizers. The Modified LeafNet model achieved the highest accuracies of 97.44% and 87.76% for the validation and testing datasets, respectively. In comparison, LeafNet obtained 88.92% and 71.84%, Xception obtained 88.64% and 71.95%, and MobileNetV2 obtained 82.10% and 67.68% for the validation and test accuracies on the same datasets, respectively. This study contributes to the development of automated disease classification systems for rice leaves, thereby leading to increased agricultural productivity and sustainability.

关键词： Plant diseases detection algorithms Deep learning Convolutional neural networks Agriculture Classification algorithms Crop yield Transfer learning Hyperparameter optimization convolutional neural networks transfer learning image classification

来源：评论

学校读者我要写书评

暂无评论

Perceptual Hashing With Deep and Texture Features

引用

IEEE MULTIMEDIA 2024年第1期31卷 65-75页

作者： Yu, Mengzhu Tang, Zhenjun Liang, Xiaoping Zhang, Xianquan Zhang, Xinpeng Guangxi Normal Univ Software Engn Guilin 541004 Peoples R China Guangxi Normal Univ Sch Comp Sci & Engn Guilin 541004 Peoples R China Fudan Univ Sch Comp Sci Shanghai 200433 Peoples R China

Image hashing is a useful technique for many multimedia systems, such as image authentication, image copy detection, tampering detection, and image quality assessment (IQA). However, most of the image hashing schemes do not make desirable performance of IQA. To tackle this, a new hashing scheme with deep and texture features is proposed for reduced-reference (RR)-IQA. In the proposed hashing, deep features are calculated from the discrete cosine transform coefficients of the three-order tensor stacked by the feature maps generated by the pretrained ResNet18. Texture features are extracted by the gray-level co-occurrence matrix in the nonsubsampled shearlet transform domain. Hash is determined by combining the quantized versions of the deep and texture features. Extensive experiments performed on open datasets indicate that the proposed perceptual hashing is superior to some baseline schemes in the performances of RR-IQA and classification.

关键词： Feature extraction Discrete cosine transforms Tensors Distortion Robustness Three-dimensional displays Sensitivity Transforms Image quality Multimedia systems Classification algorithms detection algorithms Hash functions

来源：评论

学校读者我要写书评

暂无评论

A Principal Component Analysis-Based Approach for Single Morphing Attack detection

A Principal Component Analysis-Based Approach for Single Mor...

引用

23rd IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)

作者： Dargaud, Laurine Ibsen, Mathias Tapia, Juan Busch, Christoph Tech Univ Denmark DTU Lyngby Denmark Hsch Darmstadt Hda Da Sec Biometr & Internet Secur Res Grp Darmstadt Germany

ISBN: (纸本)9798350320565

This paper proposes an explicit method for single face image morphing attack detection, using an RGB decomposition based on Principal Component Analysis from texture patterns. Handcrafted detection algorithms can be advantageous over deep learning-based methods as they constitute increased explainability, showcased in this work by visualizing relevant face areas for morphing attack detection. Such information can be relevant for deployed systems in real-world scenarios with humans in the loop. The morphing detection capability of the proposed method is evaluated extensively across three datasets and six morphing algorithms in single, cross-dataset and cross-morphed scenarios and compared to a fine-tuned MobileNetV2 architecture. The results show how single image morphing attack detection remains challenging, especially in cross-domain scenarios involving realistic diversity of morphing algorithms, including StyleGAN-based approaches. In such conditions, the proposed method can be as good or even better than the evaluated MobileNetV2 approach.

关键词： Training Learning systems Visualization Image color analysis Feature extraction Human in the loop detection algorithms

来源：评论

学校读者我要写书评

暂无评论

Compressive Sensing-Based Joint Activity and Data detection for Grant-Free Massive IoT Access

引用

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS 2022年第3期21卷 1851-1869页

作者： Mei, Yikun Gao, Zhen Wu, Yongpeng Chen, Wei Zhang, Jun Ng, Derrick Wing Kwan Di Renzo, Marco Beijing Inst Technol Sch Informat & Elect Adv Res Inst Multidisciplinary Sci Beijing 100081 Peoples R China Beijing Inst Technol Adv Res Inst Multidisciplinary Sci Beijing 100081 Peoples R China Beijing Jiaotong Univ State Key Lab Rail Traff Control & Safety Beijing 100044 Peoples R China Shanghai Jiao Tong Univ Dept Elect Engn Shanghai 200240 Peoples R China Univ New South Wales Sch Elect Engn & Telecommun Sydney NSW 2025 Australia Univ Paris Saclay Lab Signaux & Syst CNRS Cent SupelecUniv Paris Sud F-91192 Paris France

Massive machine-type communications (mMTC) are poised to provide ubiquitous connectivity for billions of Internet-of-Things (IoT) devices. However, the required low-latency massive access necessitates a paradigm shift in the design of random access schemes, which invokes a need of efficient joint activity and data detection (JADD) algorithms. By exploiting the feature of sporadic traffic in massive access, a beacon-aided slotted grant-free massive access solution is proposed. Specifically, we spread the uplink access signals in multiple subcarriers with pre-equalization processing and formulate the JADD as a multiple measurement vectors (MMV) compressive sensing problem. Moreover, to leverage the structured sparsity of uplink massive access signals among multiple time slots, we develop two computationally efficient detection algorithms, which are termed as orthogonal approximate message passing (OAMP)-MMV algorithm with simplified structure learning (SSL) and accurate structure learning (ASL). To achieve accurate detection, the expectation maximization algorithm is exploited for learning the sparsity ratio and the noise variance. To further improve the detection performance, channel coding is applied and successive interference cancellation (SIC)-based OAMP-MMV-SSL and OAMP-MMV-ASL algorithms are developed, where the likelihood ratio obtained in the soft-decision can be exploited for refining the activity identification. Finally, the state evolution of the proposed OAMP-MMV-SSL and OAMP-MMV-ASL algorithms is derived to predict the performance theoretically. Simulation results verify that the proposed solutions outperform various state-of-the-art baseline schemes, enabling low-latency random access and high-reliable massive IoT connectivity with overloading.

关键词： Inference algorithms Approximation algorithms Uplink OFDM Performance evaluation Wireless communication detection algorithms Compressive sensing grant-free massive access orthogonal approximate message passing multiple measurement vectors successive interference cancellation

来源：评论

学校读者我要写书评

暂无评论

DC Series Arc Fault Detector With Differential DWT and Variable Threshold Method for Photovoltaic Systems

引用

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 2024年第7期20卷 9343-9351页

作者： Ahn, Jae-Beom Jeong, Seung-Jae Cho, Chan-Gi Ryoo, Hong-Je Chung Ang Univ Dept Energy Syst Engn Seoul 06974 South Korea Agcy Def Dev Dept Laser Weap Syst PMO Daejeon 34186 South Korea Chung Ang Univ Sch Energy Syst Engn Seoul 06974 South Korea

A series arc fault detector (AFD) is a significant device for preventing fire hazards in photovoltaic (PV) systems. The AFD should detect a series arc quickly and accurately. However, system noise due to the components of a PV system can cause false detection of the AFD. Furthermore, as the inverter types vary according to PV systems and the irradiation changes during one day, it is difficult to develop a universal arc fault detection algorithm with an adequate arc detection criterion, and these difficulties limit the commercialization of AFDs. This study proposes an arc detection algorithm based on differential discrete wavelet transform (DWT) analysis and variable threshold method. Differential DWT analysis increases the distinction performance of inverter noise and arc noise and detects series arcs quickly and effectively by acquiring the amplified arc noise and attenuated inverter noise. Moreover, the variable threshold method updates the adequate threshold level for arc detection in real time and does not require manual tuning for systems and irradiation change. This study also proposes an AFD based on the TMS320F28335 digital signal processor to analyze noise and detect the series arc fault in real time using the proposed algorithm. The performance of the series AFD was verified via a series arc-detection test under the UL1699B test facility. Furthermore, the reliability of the arc detection algorithm and AFD was verified via the unwanted tripping test and arc detection under a real PV system.

关键词： Discrete wavelet transforms Noise Inverters Radiation effects detection algorithms History Real-time systems Arc fault detector (AFD) discrete wavelet transform (DWT) fault diagnosis photovoltaic (PV) system series arc

来源：评论

学校读者我要写书评

暂无评论

A Fast Community detection Algorithm Based on Reconstructing Signed Networks

引用

IEEE SYSTEMS JOURNAL 2022年第1期16卷 614-625页

作者： Xia, Chengyi Luo, Yongping Wang, Li Li, Hui-Jia Tianjin Univ Technol Tianjin Key Lab Intelligence Comp & Novel Softwar Engn Res Ctr Learning Based Intelligent Syst Minist Educ Tianjin 300384 Peoples R China Beijing Univ Posts & Telecommun Sch Sci Beijing 100876 Peoples R China

Signed networks depict the individual cooperative or hostile relationship in a population, which can help to deeply mine the characteristics of complex networks and predict the potential collaboration between individuals by analyzing their interaction within different groups or communities. In this article, first of all, an improved modularity function for signed networks is proposed on the basis of the existing modularity function. Then, a new community detection algorithm for signed networks has also been devised, and time complexity analysis shows that the time required for the algorithm has a linear relationship with the number of nodes in the sparse networks. Meanwhile, the affinity index that can be used to convert directed signed networks into the corresponding undirected signed networks is come up with. Finally, the current algorithm has been applied into several illustrative and realistic networks. The experimental results indicate that the number of communities given by the proposed algorithm is consistent with that of actual communities, and thus, it can be further conducive to identifying the community structure hidden within the real-world systems.

关键词： Indexes Partitioning algorithms Image edge detection Clustering algorithms Oscillators Heuristic algorithms detection algorithms Affinity index community detection community structure modularity network reconstruction signed network

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：