检索结果-内蒙古大学图书馆

2023 International Conference on Integrated Intelligence and Communication Systems, ICIICS 2023

作者： Su, Yunxia Shanghai Publishing and Printing College Shanghai China

ISBN: (纸本)9798350315455

With the popularization of the internet, people have paid more and more attention to image processing, and deep learning has become a research hotspot. Analyzing images based on deep neural networks is a very valuable and promising direction for development. This article first introduced text representation methods and two commonly used classifiers in machine learning, namely the manual annotation model method. Then, it elaborated on the effectiveness and current application status of a fusion of multi-layer feedforward technology and convolutional neural networks for edge detection, feature extraction, and other fields. Finally, a new concept was proposed, and deep learning was applied to explain the important significance of image preprocessing in digital music file classification. Afterwards, a communication field training model based on image recognition algorithms was simulated and tested for its functionality. The test results showed that the image recognition algorithm based on deep learning had good noise processing performance during the propagation process, with an image clarity rate of over 90%. Through deep learning algorithms, automatic recognition and classification of radio signals can be achieved, including determining signal type, modulation method, frequency range, and other information. This is of great significance for radio spectrum monitoring, spectrum management, and interference source localization. By using deep learning algorithms, classification models for different types of radio signals can be trained, making the recognition of radio spectrum more accurate and reliable. © 2023 IEEE.

关键词： communication field deep learning image recognition algorithms neural networks

来源：评论

学校读者我要写书评

暂无评论

Pavement damage detection and evaluation based on UAV image and improved AHP

引用

CASE STUDIES IN CONSTRUCTION MATERIALS 2025年 22卷

作者： Zuo, Dunren Shao, Lei Chen, Bohan Tan, Jingxin Yang, Jiahao Chongqing Univ Sch Resources & Safety Engn Chongqing Peoples R China Chongqing Univ Coll Aerosp Engn Chongqing Peoples R China Chongqing Jiaotong Univ Sch Aeronaut Chongqing Peoples R China

As a crucial component of road infrastructure, the pavement significantly impacts the comfort of vehicle and pedestrian traffic. Various factors such as traffic flow, weather conditions, and material quality can lead to pavement damage. This paper utilizes UAVs to collect road damage images and establishes a classification and evaluation model by using image recognition algorithms and the analytic hierarchy process (AHP) theory. To verify the effectiveness of the model, a road in Chongqing was selected as the experimental object. This model can precisely and effectively distinguish different types of pavement damage, such as ruts, grooves, and so on. The results show that the classification accuracy of the model is 92.5 %, which has good detection performance compared with other models. In the research, it was found that the contrast of rut damage images varies greatly, ranging from 0.08 to 0.32. This difference may be due to the larger damage area present in the rut image compared to the other images. Considering groove and crack features are similar, there is a 66.7 % chance of identifying one as the other.

关键词： Pavement damage AHP Gray level co-occurrence matrix UAV image recognition algorithms

来源：评论

学校读者我要写书评

暂无评论

Descriptive image Analysis: Part III. Multilevel Model for algorithms and Initial Data Combining in Pattern recognition

引用

PATTERN recognition AND image ANALYSIS 2020年第3期30卷 328-341页

作者： Gurevich, I. B. Yashina, V. V. Russian Acad Sci Fed Res Ctr Comp Sci & Control Bldg 244 Vavilov Str Moscow 119333 Russia

This is the third article in a series of publications devoted to the current state and future prospects of Descriptive image Analysis (DIA), one of the leading and intensively developing branches of modern mathematical theory of image analysis. The fundamental problem of computer science touched by the article is the automation of extracting from images of information necessary for intellectual decision-making. A new class of models for the image analysis and recognition process and its constituent procedures is introduced and described - a multilevel model of image analysis and recognition procedures (MMCAI) - which is based on the joint use of methods of combining algorithms and methods of combining fragmentary initial data - partial descriptions of the object of analysis and recognition - an image. The architecture, functionality, limitations, and characteristics of the MMCAI are justified and defined. The main properties of the MMCAI class are as follows: (a) combining the fragments of the initial data and their representations and combining algorithms at all levels of image analysis and recognition processes;(b) the use of multialgorithmic schemes in the image analysis and recognition process;and (c) the use of dual representations of images as input data for the analysis and recognition algorithms. The problems arising in the development of the MMCAI are closely related to the development of the following areas of the modern mathematical theory of image analysis: (a) algebraization of image analysis;(b) image recognition algorithms accepting spatial information as input data;(c) multiple classifiers (MACs). A new class of models for image analysis is introduced in order to provide the following possibilities: (a) standardization, modeling, and optimization of Descriptive Algorithmic Schemes (DAS) that form the brainware of the MMCAI and processing heterogeneous ill-structured information - dual representations - spatial, symbolic, and numerical representations

关键词： image analysis descriptive image analysis mathematical theory of image analysis models of image analysis procedures image representations image models descriptive image representations descriptive image models dual representation of images combinatorial structures of local neighborhoods multiple classifiers support sets image recognition algorithms

来源：评论

学校读者我要写书评

暂无评论

Research on intelligent endpoint determination in copper converting based on machine learning

引用

CANADIAN METALLURGICAL QUARTERLY 2025年

作者： Zhao, Bi Hao, Haibin Xu, Jianxin Chen, Enlin Lu, Yongjie Wang, Hua Kunming Univ Sci & Technol Fac Met & Energy Engn Kunming 650093 Peoples R China Kunming Univ Sci & Technol State Key Lab Complex Nonferrous Met Resources Cle Kunming 650093 Peoples R China Liangshan Min Ltd Co Liangshan Peoples R China

Copper converters are pivotal in the copper smelting process, playing a critical role in producing high-quality copper. Accurate endpoint determination in these converters is essential for maintaining copper quality and enhancing smelting efficiency. This study addresses challenges such as slag formation, the rudimentary level of intelligence in endpoint prediction, and low accuracy rates in converters. We propose a dynamic, multitask model that integrates converter operational conditions, data collection on smelting parameters, and dynamic endpoint prediction to intelligently predict the termination of copper and slag production. A multitask dynamic model based on a Deep Convolutional Neural Network (DCNN) was developed through the collection and analysis of data during the smelting process. This model predicts and optimises the smelting process within copper converters. Our results demonstrate that the model achieved an average prediction error of 1.9% for slag formation and 1.8% for copper production endpoint prediction. Furthermore, it attained hit rates of 97.8% and 98.3% for slag formation and copper production endpoints, respectively. Consequently, this model significantly enhances the accuracy of endpoint determination in slag and copper production, thus improving the efficiency and quality of output from copper converters. Les convertisseurs de cuivre sont essentiels dans le proc & eacute;d & eacute;de fusion du cuivre, jouant un r & ocirc;le critique dans la production de cuivre de haute qualit & eacute;. Une d & eacute;termination pr & eacute;cise du point final dans ces convertisseurs est essentielle pour maintenir la qualit & eacute;du cuivre et am & eacute;liorer l'efficacit & eacute;de la fusion. Cette & eacute;tude r & eacute;pond aux d & eacute;fis tels que la formation de scories, le niveau rudimentaire d'intelligence dans la pr & eacute;diction du point final et les faibles taux de pr & eacute;cision des convertisseurs. Nous proposons un mod & egr

关键词： Copper converting neural networks image recognition algorithms endpoint determination multitask learning

来源：评论

学校读者我要写书评

暂无评论

High precision angular displacement measurement based on self-correcting error compensation of three image sensors

引用

APPLIED OPTICS 2022年第1期61卷 287-293页

作者： Yu, Hai Wan, Qiuhua Lu, Xinran Zhao, Changhai Liang, Lihui Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys Changchun 130033 Peoples R China

As the feedback link of a numerical control system, the measurement accuracy of absolute angular displacement measurements directly affects the control performance of a numerical control system. In previous research, angular displacement measurements based on dual image sensors can achieve higher measurement performance. However, the elimination of harmonic error by the dual image sensor is still limited. For this reason, this paper proposes an image-type angular displacement measurement method based on self-correction error compensation of three image sensors. First, the mathematical model of harmonic error is established, and the shortcomings of using dual image sensors to compensate the error are analyzed. Then, a high precision angular displacement measurement method based on three image sensors is proposed. Fmally, the self-correction error compensation method of three image sensors is applied to the angular displacement measurement system, and the measurement performance is verified. The experimental results show that a measurement accuracy of 1.76 '' can be achieved on the circular grating with a diameter of 96 mm. In contrast, the dual image sensor can only achieve a measurement accuracy of 2.88 ''. It is concluded that the odd number of image sensors can achieve higher measurement accuracy than the even number. This research lays a foundation for the realization of high precision image angular displacement measurement. (C) 2021 Optica Publishing Group

关键词： Charge coupled devices Digital image processing image recognition algorithms image sensors Precision metrology Spatial resolution

来源：评论

学校读者我要写书评

暂无评论

image processing algorithms in the assessment of grain damage degree

引用

BIO-algorithms AND MED-SYSTEMS 2022年第1期18卷 17-28页

作者： Maliga, Wojciech Dudzinski, Wlodzimierz Labowska, Magdalena Detyna, Jerzy Lopusiewicz, Marcin Bujak, Henryk Wroclaw Univ Sci & Technol Dept Mech Mat & Biomed Engn Smoluchowskiego 25 PL-50372 Wroclaw Poland Wroclaw Univ Sci & Technol Dept Mat Sci Welding & Strength Mat Wroclaw Poland Wroclaw Univ Environm & Life Sci Dept Genet Plant Breeding & Seed Prod Wroclaw Poland Res Ctr Cultivar Testing Slupia Wielka Poland

Objectives: The paper presents preliminary results on the assessment of algorithms used in image processing of the grain damage degree. The purpose of the work is developing a tool allowing to analyse sample cross-sections of rye germs. Methods: The analysis of the grain cross-sections was carried out on the basis of a series their photos taken at equal time intervals at a set depth. The cross-sections will be used to create additional virtual cross-sections allowing to analyse the whole sample volume. The ultimate plan is to generate two cross-sections perpendicular to each other. Based on volumetric data read from the sample section, a three-dimensional model of an object will be generated. Results: The analysis of model surface will allowed us to detect possible grain damage. The developed method of preparing the research material and the proprietary application allowed for the identification of internal defects in the biological material (cereal grains). Conclusions: The presented methodology may be used in the agri-food industry in the future. However, much research remains to be done. These works should primarily aim at significantly reducing the time-consuming nature of individual stages, as well as improving the quality of the reconstructed image.

关键词： 3D surface models grain damage image recognition algorithms seeds infection surface mapping

来源：评论

学校读者我要写书评

暂无评论

AI-Enhanced Unmanned Aerial Vehicles for Search and Rescue Operations 5

AI-Enhanced Unmanned Aerial Vehicles for Search and Rescue O...

引用

5th International Conference on Innovative Trends in Information Technology (ICITIIT)

作者： Farsath, Rashida K. Jitha, K. Marwan, Mohammed V. K. Jouhar, Muhammed Ali A. Farseen, Muhammed K. P. Musrifa, K. A. MEA Engn Coll Dept Comp Sci & Engn Malappuram India

ISBN: (纸本)9798350386813;9798350386820

This paper introduces a cutting-edge AI-empowered Unmanned Aerial Vehicle (UAV) system, enriched with stateof-the-art sensor technology, advanced image recognition algorithms, and autonomous navigation capabilities. The system represents a transformative approach to search and rescue operations, offering unparalleled precision and rapid response times. Our methodology encompasses multifaceted data collection techniques, including surveys, interviews, data mining, Internet of Things (IoT) sensors, and sophisticated video analytics. Machine learning and deep learning models are then applied to process and analyze this data, enabling real-time image recognition for precise target identification. The system's AI-driven autonomous navigation algorithms optimize mission planning, resulting in significantly reduced response times and heightened mission success rates. Extensive real-world tests and simulations validate the exceptional performance of the proposed AI-empowered UAV system. These tests underscore its capacity to expedite emergency response efforts in dynamic and challenging environments. In parallel, this paper addresses critical ethical considerations, ememphasizing responsible data handling practices, and robust security measures to ensure the system's integrity in sensitive contexts. As exemplified through a compelling case study of successful rescue operations, this technology represents a groundbreaking advancement in the field. By bridging the gap between cutting- edge technology and life-saving applications, it holds the potential to redefine the landscape of search and rescue missions, ushering in an era of heightened efficiency, precision, and impact.

关键词： Artificial Intelligence Unmanned Aerial Vehicles Sensor Technology image recognition algorithms Autonomous Navigation Search and Rescue Operations Emergency Response Human-Machine Collaboration Drones Machine Learning Deep Learning Data Privacy Ethics IOT Sensors Video analytic IoT Deep Learning Random Forest Restricted Boltzmann Network Auto encoder Intrusion Detection System Smart City

来源：评论

学校读者我要写书评

暂无评论

Multialgorithmic Hierarchical image Analysis System: Architecture and Analysis Model

引用

PATTERN recognition AND image ANALYSIS 2024年第4期34卷 959-965页

作者： Gurevich, I. B. Yashina, V. V. Russian Acad Sci Fed Res Ctr Comp Sci & Control Moscow 119333 Russia

The article gives a detailed description of one of the basic models of descriptive image analysis, characterizing the architecture and structure of the image recognition process: a multilevel model of image analysis and recognition procedures with joint use of methods for combining algorithms and fragmentary initial data-partial descriptions of the object of analysis and image recognition. The architecture, functional capabilities, limitations, and characteristics of a multilevel model for combining algorithms and initial data in image recognition are substantiated and defined.

关键词： image analysis descriptive image analysis mathematical theory of image analysis automated information extraction from images image recognition descriptive algorithmic schemes models of image analysis procedures image representations image models image recognition algorithms mathematical structures formal systems multialgorithmic classifiers dual image representations

来源：评论

学校读者我要写书评

暂无评论

Rapid identification of the variety of maize seeds based on near-infrared spectroscopy coupled with locally linear embedding

引用

APPLIED OPTICS 2022年第7期61卷 1704-1710页

作者： Liu, Shu Chen, Zhengguang Jiao, Feng Heilongjiang Bayi Agr Univ Coll Informat & Elect Engn Daqing 163319 Peoples R China Heilongjiang Bayi Agr Univ Coll Agr Daqing 163319 Peoples R China

Maize is the main cereal crop in China. In the process of maize planting, the selection of suitable maize varieties is an important link to achieving a high yield. Because the appearance of maize seeds is very similar, it is difficult to accurately identify their species with the naked eye. In order to realize the rapid identification of different varieties of maize seeds, this paper proposes a rapid identification method of maize varieties based on near-infrared (NIR) spectroscopy coupled with locally linear embedding (LLE) and a support vector machine (SVM). The MR data, preprocessed by multiple scattering correction (MSC), were dimensionally reduced by LLE, a principal component analysis (PCA), and isometric mapping (Isomap), and combined with SVM to establish a maize variety identification model. The results show that the LLE-SVM model has the best performance, whose classification accuracy and kappa coefficient of the test set can reach 100% and 1.00. The classification accuracy and kappa coefficient of the LLE-SVM model are better than the PCA-SVM model and Isomap-SVM model. Therefore, LLE can reduce the complexity of the model and improve the accuracy of the model. It can be used for the rapid identification of maize varieties and provide a new idea for the classification and identification of other agricultural products. (C) 2022 Optica Publishing Group

关键词： Face recognition Fourier transforms Genetic algorithms image recognition algorithms Multiple scattering Near infrared spectroscopy

来源：评论

学校读者我要写书评

暂无评论

High-precision displacement measurement algorithm based on a depth fusion of grating projection pattern

引用

APPLIED OPTICS 2022年第4期61卷 1049-1056页

作者： Yu, Hai Wan, Qiuhua Lu, Xinran Du, Yingcai Liang, Lihui Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys Changchun 130033 Peoples R China

Displacement measurement based on an image recognition algorithm has more advantages in the field of displacement measurement. Due to the uncertainty of image location, the accuracy of the displacement measurement algorithm studied in the previous stage is limited. To improve the accuracy of image-type displacement measurement, a high-precision displacement measurement algorithm based on depth fusion of the grating projection pattern is proposed. First, a line-scan image sensor and parallel light source are placed on both sides of the calibration grating. The marking pattern on the calibration grating is projected onto the image sensor through illumination of the parallel light source to realize the projection imaging. Then, the centroid algorithm is used to calculate the centroid of each line in the collected image, and the mean value of the centroid group is determined. Finally, a linear function is used to fit multiple centroids to calculate the endpoints of the subdivision region, and the subdivision operation is carried out according to a ratio algorithm. To test the performance of the proposed algorithm, a linear displacement measurement device with a 250 mm range and an angular displacement measurement device with a 100 mm diameter grating are both used for experiments. After testing, the linear measurement accuracy of the proposed algorithm is improved from 4.63 to 1.26 mu m, and the angular displacement measurement accuracy is enhanced from 8.87 to 2.88 ''. The research in this paper provides a theoretical basis for high-performance displacement measurement. (C) 2022 Optica Publishing Group

关键词： Digital image processing image detection systems image processing algorithms image recognition algorithms image sensors Spatial resolution

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：