检索结果-内蒙古大学图书馆

Data Augmentation and Intelligent recognition in Pavement Texture Using a Deep Learning

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS 2022年第12期23卷 25427-25436页

作者： Chen, Ning Xu, Zijin Liu, Zhuo Chen, Yihan Miao, Yinghao Li, Qiuhan Hou, Yue Wang, Linbing Beijing Univ Technol Beijing Key Lab Traff Engn Beijing 100124 Chaoyang Peoples R China Toyota Transportat Res Inst Toyota Aichi 4710024 Japan Southeast Univ Sch Transportat Nanjing 211189 Peoples R China Univ Sci & Technol Beijing Natl Ctr Mat Serv Safety Beijing 100083 Peoples R China Beijing Jiaotong Univ Key Lab Transport Ind Big Data Applicat Technol C Beijing 100091 Peoples R China Virginia Tech Dept Civil & Environm Engn Blacksburg VA 24061 USA

As one of the most important properties of road, the adaption to roads with different macro-textures may significantly affect the autonomous driving technologies since road texture directly affects the skidding resistance and tire noise. Therefore, it is of great significance to detect and analyze the road macro texture with respect to different pavement types and service conditions. Generally, transportation engineers may face problems such as small dataset size, unbalanced dataset, etc. To solve these problems, this study aims to recognize the pavement texture using the deep learning approaches. The pavement texture data was first visualized using image processing methods, and then augmented using the traditional methods as well as a deep learning approach, i.e. Generative Adversarial Network (GAN) model. The Random Forest (RF) algorithm and the DenseNet network were both employed, where the overall classification accuracy of the original dataset was 50% and 59%, respectively, and the accuracy of the data augmented by the traditional methods was 58% and 70%, respectively. Test results show that, after 250,000 generations of training, GAN model was able to generate new pavement texture images with high quality, and the classification accuracy on the test dataset using DenseNet improved to 82%. It was discovered that the deep learning methods had a better performance for pavement texture recognition than manual classification and traditional machine learning methods. Furthermore, it was also found that adding noise in the original datasets as an augmentation method had a negative impact on the classification accuracy.

关键词： Deep learning Training Roads Feature extraction Random forests Generative adversarial networks Data models Data augmentation pavement texture recognition and classification deep learning autonomous driving

来源：评论

学校读者我要写书评

暂无评论

Industrial Sheet Image Processing Technology Based on Machine Vision

Industrial Sheet Image Processing Technology Based on Machin...

引用

2024 International Conference on Artificial Intelligence and Machine Learning Research, CAIMLR 2024

作者： Guo, Wenxin Cheng, Xuefeng Dai, Xueling School of Big Data and Information Industry Chongqing City Management College Chongqing 401331 China

ISBN: (纸本)9781510691797

This paper designs and implements an efficient and accurate machine vision-based stator sheet image processing algorithm to address the field of precision stacking of industrial sheets. By comprehensively applying key technologies such as image processing, feature extraction, and recognition classification, the algorithm first performs preprocessing steps such as downsampling, binarization, and morphology on the original image, effectively eliminating interference information, enhancing target features, significantly improving image processing speed and efficiency, and meeting the real-time requirements of the production line. Then, a radial line search method is used to accurately locate the edge points of the sheet, enabling precise recognition and classification of stator sheets in both forward and reverse directions, ensuring high accuracy and reliability of the recognition process. The results of this study can be applied in industrial practice, driving automation and intelligence, improving production efficiency, reducing labor costs, and significantly enhancing product quality. © 2025 SPIE.

关键词： Feature Extraction Image Processing Industrial Automation Machine Vision recognition and classification

来源：评论

学校读者我要写书评

暂无评论

Revealing the real-world applicable setting of online continual learning 24

Revealing the real-world applicable setting of online contin...

引用

IEEE 24th International Workshop on Multimedia Signal Processing (MMSP)

作者： Xu, Zhenbo Hu, Haimiao Liu, Liu Beihang Univ Hangzhou Innovat Inst Hangzhou Peoples R China ShiFang Technol Inc Hangzhou Peoples R China

ISBN: (数字)9781665471893

ISBN: (纸本)9781665471893

The motivation of online continual learning (CL) is training agents to learn from an infinite stream of data and quickly accommodate changes in the data distribution. However, current online CL datasets are synthesized by common classification datasets by splitting all classes into disjoint tasks where disjoint task streams have little temporal relations, resulting in a CL setting far from realistic. In this paper, we ask two questions: (i) What are the characteristics of real-world CL scenarios? (ii) How existing methods perform on real-world CL scenarios? To answer the first question, we propose the first realistic CL setting coined instance-based continual learning (IBCL). IBCL has no task or class boundaries and requires algorithms to predict and learn from instance streams simultaneously. The life cycles of classes under IBCL are dynamic and instances belonging to the same class might evolve over time. For each sequentially arrival instance, algorithms are required to give the recognition result and then perform changes based on its label. No additional training resource are available except for the instance stream in evaluation. To answer the second question, on CORe50 and mini-ImageNet, we compare current online CL methods under the IBCL setting with both the traditional ResNetl8 backbone as well as the recent transformer-based backbone ViT on the IBCL setting. Three aspects including the recognition performance, the latency, and the memory usage of current methods are analyzed. Experiment results show that current online CL methods perform poorly in the real CL scenarios, and methods using the transformer-based backbone perform better than the CNN-based counterparts.

关键词： Online Continual Learning recognition and classification

来源：评论

学校读者我要写书评

暂无评论

On line identification method of forming defects of small module plastic gears 10

On line identification method of forming defects of small mo...

引用

10th International Symposium on Precision Mechanical Measurements

作者： Shi Zhaoyao Fang Yiming Song Huixu Beijing Univ Technol Fac Mat & Mfg Beijing Engn Res Ctr Precis Measurement Technol & Beijing 100124 Peoples R China

ISBN: (数字)9781510649941

ISBN: (纸本)9781510649941;9781510649934

Due to its anisotropic shrinkage characteristics and the instability of forming process, small module plastic gears are easy to produce complex appearance defects after forming. The traditional manual detection can not accurately identify all kinds of defects, let alone complete the full inspection and statistical analysis. Therefore, a prototype of small module plastic gear on-line detection and sorting system based on machine vision is developed. The system has the function of on-line classification and identification of plastic gear surface defects, and can realize full inspection. The watershed algorithm based on local extreme value and Canny edge extraction algorithm can accurately extract the gear defect area, mark the defect location and count the defect characteristics, complete the determination of defects such as size out of tolerance, shrinkage, burr, surface foreign matter and incompletely filled part in 0.3s, and display the defect location through the human-computer interaction interface, with a discrimination accuracy of 95%.

关键词： machine vision plastic gear molding defects recognition and classification

来源：评论

学校读者我要写书评

暂无评论

Identification And classification of Rail Damage Based On Ultrasonic Echo Signals 39

Identification And Classification of Rail Damage Based On Ul...

引用

39th Chinese Control Conference (CCC)

作者： Li Yujun Yao Fengtao Jiao Shangbin Huang Weichao Zhang Qing Xian Univ Technol Sch Automat & Informat Engn Xian 710048 Peoples R China

ISBN: (数字)9789881563903

ISBN: (纸本)9789881563903

Rail ultrasonic flaw detection is an important method for railway operation and maintenance. The accuracy of damage detection directly affects the safety of railway transportation systems. Rail damage detection, identification and classification are the key to ultrasonic flaw detection systems. At present, most of the researches focus on the local damage of rails, such as rail head nuclear damage, rail waist cracks, or through holes at the bottom of the rail. The classification and identification of typical damages of the entire rail have not been systematically studied. In this paper six features are extracted by principal component analysis(PCA), they are the ratio of the energy of the signal processed by wavelet decomposition, the area surrounded by the 100-150 sampling points and the horizontal axis, the maximum amplitude and its position of the 100-150 sampling points, and the maximum amplitude in the 240-300 sampling points, and the position in the 300-400 sampling points, which are used to present the characteristics of the rail damage. Least square support vector machine (LSSVM) is introduced to create the classification analysis models, the eigenvalues extracted by PCA are the input of the models and any rail damage is the expectation output. In order to improve the accuracy of the classification models, particle swarm optimization (PSO) algorithm is introduced to optimize the parameters of the model, seven classification analysis models of the rail damage are built. Experiments results show that the typical damage in the standard test rail can be accurately identified and classified, the classification recognition accuracy is about 92%. This method can recognize the typical rail damage in the national standard. It has definite practical application value in the field of rail inspection.

关键词： rail damage LSSVM PCA PSO recognition and classification

来源：评论

学校读者我要写书评

暂无评论

recognition and classification for Vision Navigation Vehicle in Agricultural Environment based on MSBN

Recognition and Classification for Vision Navigation Vehicle...

引用

24th Chinese Control and Decision Conference (CCDC)

作者： Guo, Wenqiang Zhu, Zoe Hou, Yongyan Fu, Ju Shaanxi Univ Sci & Tech Sch Elect & Informat Engn Xian 710021 Peoples R China Univ Guelph Dept Comp Informat Sci Guelph ON N1G2W1 Canada Shaanxi Univ Sci Sch Elect & Informat Engn Xian 710021 Peoples R China Shaanxi Univ Sci Sch Elect & Informat Engn Xian 710021 Peoples R China

ISBN: (纸本)9781457720727

To solve the uncertain problem in the agricultural environment recognition and classification for vehicles, an environment recognition algorithm for vehicles based on inference in the multiply sectioned Bayesian network (MSBN) is proposed. This method represents multiple image sensor systems into sub-Bayesian networks in the MSBN. With the existing local and global exact inference algorithm in MSBN, the presented method can improve the recognition performance by fusion multi-source partial observation evidences from sub-Bayesian networks via their effective updated belief communication among the subnets. Experimental results illustrate that this MSBN-based agricultural environment recognition and classification approach for vehicles' navigation system can provide more accurate results than the existing Bayesian network method, with the attractive handling with uncertain and incomplete observation in the single sensor system.

关键词： recognition and classification Agricultural Environment Multiply Sectioned Bayesian Network (MSBN) Inference

来源：评论

学校读者我要写书评

暂无评论

Detection and classification of glass defects based on machine vision 3

Detection and classification of glass defects based on machi...

引用

Conference on Applied Optical Metrology III

作者： Jiang, Jiabin Xiao, Xiang Feng, Guohua Lu, ZiChen Yang, Yongying Zhejiang Univ Coll Opt Sci & Engn State Key Lab Modern Opt Instrumentat Hangzhou 310027 Peoples R China

ISBN: (数字)9781510628984

ISBN: (纸本)9781510628984;9781510628977

Surface defects inspection is a critical part in manufacturing of mobile phone cover glass. Considering the defects in the optical elements surface caused by the imperfection of manufacturing technique, the classification and the position information should be carried out for the necessary repairing process. The traditional manual inspection method is always labor-consuming and inefficient. Surface defects digital evaluation system based on machine vision draws much attention in the recent years. This paper proposed algorithms and applications for the detection task with higher efficiency and reliability comparing to the manual inspection. The detection method is based on machine vision and machine learning techniques. The images of optical elements surface are captured by line-scanning cameras, with the imaging systems of dark-field, bright-field and transmission-field. Only one image system is not enough to detect all kind of defects like scratch, bubble, crack of glass and edge chipping etc. The position information and category of defects are obtained based on image processing technique. The defective area was calculated by image filtering algorithm, the feature selection techniques based on segmentation methods are explored and the feature vector can be extracted before the next step of classification with Support Vector Machine (SVM) technique. Verified by the experiments, the results reveal this method has good performance and is very suitable for recognition and classification of glass defects.

关键词： glass defects machine vision imaging system recognition and classification

来源：评论

学校读者我要写书评

暂无评论

Identification And classification of Rail Damage Based On Ultrasonic Echo Signals

Identification And Classification of Rail Damage Based On Ul...

引用

第三十九届中国控制会议

作者： Li Yujun Yao Fengtao Jiao Shangbin Huang Weichao Zhang Qing School of Automation and Information Engineering Xi’an University of Technology

Rail ultrasonic flaw detection is an important method for railway operation and maintenance. The accuracy of damage detection directly affects the safety of railway transportation systems. Rail damage detection, identification and classification are the key to ultrasonic flaw detection systems. At present, most of the researches focus on the local damage of rails, such as rail head nuclear damage, rail waist cracks, or through holes at the bottom of the rail. The classification and identification of typical damages of the entire rail have not been systematically studied. In this paper six features are extracted by principal component analysis（PCA）, they are the ratio of the energy of the signal processed by wavelet decomposition, the area surrounded by the 100-150 sampling points and the horizontal axis, the maximum amplitude and its position of the 100-150 sampling points, and the maximum amplitude in the 240-300 sampling points, and the position in the 300-400 sampling points, which are used to present the characteristics of the rail damage. Least square support vector machine（LSSVM） is introduced to create the classification analysis models, the eigenvalues extracted by PCA are the input of the models and any rail damage is the expectation output. In order to improve the accuracy of the classification models, particle swarm optimization（PSO） algorithm is introduced to optimize the parameters of the model, seven classification analysis models of the rail damage are built. Experiments results show that the typical damage in the standard test rail can be accurately identified and classified, the classification recognition accuracy is about 92%. This method can recognize the typical rail damage in the national standard. It has definite practical application value in the field of rail inspection.

关键词： rail damage LSSVM PCA PSO recognition and classification

来源：评论

学校读者我要写书评

暂无评论

A Survey on Acoustic Signature recognition and classification Techniques For Persistent Surveillance Systems

A Survey on Acoustic Signature Recognition and Classificatio...

引用

Conference on Signal Processing, Sensor Fusion, and Target recognition XXI

作者： Alkilani, Amjad Shirkhodaie, Amir Tennessee State Univ Dept Mech & Mfg Engn Ctr Excellence Battlefield Sensor Fus Nashville TN 37209 USA

ISBN: (纸本)9780819490704

Application of acoustic sensors in Persistent Surveillance Systems (PSS) has received considerable attention over the last two decades because they can be rapidly deployed and have low cost. Conventional utilization of acoustic sensors in PSS spans a wide range of applications including: vehicle classification, target tracking, activity understanding, speech recognition, shooter detection, etc. This paper presents a current survey of physics-based acoustic signature classification techniques for outdoor sounds recognition and understanding. Particularly, this paper focuses on taxonomy and ontology of acoustic signatures resulted from group activities. The taxonomy and supportive ontology considered include: human-vehicle, human-objects, and human-human interactions. This paper, in particular, exploits applicability of several spectral analysis techniques as a means to maximize likelihood of correct acoustic source detection, recognition, and discrimination. Spectral analysis techniques based on Fast Fourier Transform, Discrete Wavelet Transform, and Short Time Fourier Transform are considered for extraction of features from acoustic sources. In addition, comprehensive overviews of most current research activities related to scope of this work are presented with their applications. Furthermore, future potential direction of research in this area is discussed for improvement of acoustic signature recognition and classification technology suitable for PSS applications.

关键词： Acoustic Signature recognition and classification Spectral Analysis Semantic Labeling Persistent Surveillance System Taxonomy and Ontology

来源：评论

学校读者我要写书评

暂无评论

recognition and classification for Vision Navigation Vehicle in Agricultural Environment based on MSBN

Recognition and Classification for Vision Navigation Vehicle...

引用

The 24th Chinese Control and Decision Conference (第24届中国控制与决策学术年会 2012 CCDC)

作者： Wenqiang Guo Zoe Zhu Ju Fu Yongyan Hou School of Electrical and Information Engineering Shaanxi Univ. of Sci. and Tech. Xi'an China 710 Department of Computing Information Science University of GuelphCanadaN1G 2W1 School of Electrical and Information Engineering Shaanxi Univ. of Sci. and Tech. Xi'an China 710

To solve the uncertain problem in the agricultural environment recognition and classification for vehicles, an environment recognition algorithm for vehicles based on inference in the multiply sectioned Bayesian network (MSBN) is proposed. This method represents multiple image sensor systems into sub-Bayesian networks in the MSBN. With the existing local and global exact inference algorithm in MSBN, the presented method can improve the recognition performance by fusion multi-source partial observation evidences from sub-Bayesian networks via their effective updated belief communication among the subnets. Experimental results illustrate that this MSBN-based agricultural environment recognition and classification approach for vehicles’ navigation system can provide more accurate results than the existing Bayesian network method, with the attractive handling with uncertain and incomplete observation in the single sensor system.

关键词： recognition and classification Agricultural Environment Multiply Sectioned Bayesian Network (MSBN) Inference

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：