检索结果-内蒙古大学图书馆

Research on Citrus Leaf Disease Recognition Using Class-Agnostic Contrastive Learning and Supervised Organizational Mapping

引用

IEEE Access 2025年 13卷 101592-101608页

作者： Xiao, Wen Shang, Jinxia Li, Fan Ao, Ou Wang, Xin Tian, Shiyi Kunming University of Science and Technology Faculty of Information Engineering and Automation Kunming650500 China Kunming University of Science and Technology Yunnan Key Lab of Computer Technology Applications Kunming650500 China

Early detection of plant diseases is crucial for enhancing agricultural productivity and ensuring crop protection. While computer vision offers scalable alternatives to manual inspection, existing methods face two understudied challenges: (1) performance degradation in cluttered real-world conditions, and (2) ineffective feature separation for randomly distributed lesions with diverse symptoms. Notably, current research predominantly relies on controlled-environment datasets, lacking publicly accessible benchmarks with authentic field complexities. To address these, we present the first large-scale citrus leaf disease dataset (5,763 images, 6 categories) capturing complex field conditions including varying illumination, occlusion, multi-scale lesions, and deceptive backgrounds. Our key methodological innovations include a three-stage weakly supervised framework that systematically resolves background interference and feature redundancy: (i) Class-agnostic contrastive localization enabling precise lesion mapping without region annotations, (ii) Stable feature initialization through region-guided classification training, (iii) Sparse orthogonal mapping with experimentally-determined nodes introducing orthogonality constraints to enhance feature separation. Through feature alignment and feature fusion, an accuracy of 97.1% is achieved (+2.4% vs ResNet-18). Evaluations on citrus leaf and public datasets demonstrate superior accuracy and interpretable lesion localization, advancing plant disease diagnosis. © 2013 IEEE.

关键词： Citrus fruits

来源：评论

学校读者我要写书评

暂无评论

Intelligent Algorithm-based Model for Predicting International Chinese Teachers' Continuous Willingness to Teach Online 3

Intelligent Algorithm-based Model for Predicting Internation...

引用

3rd IEEE International Conference on Electronic Communications, Internet of Things and Big Data, ICEIB 2023

作者： Yang, Liqing Cui, Chaowei Wang, Tianyu College of International Chinese Language Education Yunnan Normal University Kunming China College of Information Engineering and Automation Kunming University of Science and Technology Kunming China

ISBN: (纸本)9798350333862

The purpose of this study was to assess international Chinese teachers' continuous willingness to teach online. To ensure an accurate and objective evaluation of their sustained willingness to teach online, a data mining approach was employed. The research process involved identifying indicators of teachers' digital competence, constructing a model, and validating the model. We used the TAM model to create the International Chinese Teachers TAM model, which consisted of 14 indicators across 4 dimensions. Our approach involved implementing a convolutional neural network, k-clustering, and BP neural network algorithms to construct a prediction model. The results of multiple experiments demonstrated the stability and credibility of the model. This model can be used to analyze the trend of international Chinese teachers' willingness to continuously teach online, while also providing valuable data to support the development of online teaching. The research results are innovative due to the creation of a prediction model that can accurately forecast teachers' persistent willingness to teach online. © 2023 IEEE.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Hot-spot Dynamic Temperature Rise of Oil-immersed Transformer through FBG Based Multi-point Sensing System

引用

IEEE Sensors Journal 2025年

作者： Luo, Chuan Li, Zizhou Yang, Bo Zhao, Zhengang Li, Chuan Kunming University of Science and Technology Faculty of Information Engineering and Automation Kunming650500 China Kunming University of Science and Technology Kunming650500 China

The hot spot temperature rise in oil-immersed transformers directly influences the transformer’s load capacity and thermal aging rate. However, accurately measuring hot spot temperature rise using optical fiber sensors presents a significant engineering challenge, given the transformer’s load fluctuations, complex internal oil flow disturbances, and stress variations. In this paper, we propose a multi-point sensing system based on FBGs to accurately measure the location and temperature increase of winding hot spots under variable operating conditions. Specifically, two stressinsensitive FBG temperature sensors, made entirely of dielectric materials, are designed and fabricated, demonstrating linearity of 0.99% to 1.65%, repeatability of 1.07% to 1.45% and a response time of approximately 10 s. To build a comprehensive monitoring system, multiple FBG temperature sensors are installed on the iron-core surface, within the windings, and in the vertical oil channel. During the temperature rise test, the spatial and temporal variations of temperature inside the transformer are obtained. The hot spot winding temperature rise reaches 62.9°C, and the top oil temperature rise is 39.6°C. The steady-state hot spot is concentrated in the 10th to 12th sections of the winding. The system achieves 65 mm spatial resolution vertically and enhances to 30 mm in critical hotspots. This study grasps the dynamic temperature rise rule of oil-immersed transformer hot spots under load fluctuations, providing data support for optimizing the transformer’s internal structure and regulating its load. © 2001-2012 IEEE.

关键词： Optical fibers

来源：评论

学校读者我要写书评

暂无评论

Soft Sensing Method Based on Online Dynamic Clustering and Self-labeling 11

Soft Sensing Method Based on Online Dynamic Clustering and S...

引用

11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022

作者： Wang, Yuechen Jin, Huaiping Yang, Biao Liu, Haipeng Kunming University of Science and Technology Faculty of Information Engineering and Automation Department of Automation Kunming650500 China Kunming University of Science and Technology Yunnan Key Laboratory of Artificial Intelligence Kunming650500 China

ISBN: (数字)9781665496759

ISBN: (纸本)9781665496759

Traditional soft sensing methods encounter low prediction accuracy owing to complex process characteristics, such as time-varying and nonlinear behaviors, as well as the scarcity of labeled data. A soft sensing method based on online dynamic clustering and self-labeling (ODCSL) is proposed for data stream mining of industrial processes. This method identifies the process state of the data stream through online clustering analysis, and then adaptively establishes a local Gaussian process regression (GPR) for online prediction. During each prediction run, self-labeling is performed on the recently unlabeled data to provide good pseudo-labeled data, which enhances the prediction of local GPR models. Moreover, when a new sample is available, the clustering results and the GPR models are updated. The effectiveness of the proposed ODCSL method is verified using an industrial fermentation process. © 2022 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Maximum Power Point Tracking technology for PV Systems: Current Status and Perspectives

引用

Energy engineering 2024年第8期121卷 2009-2022页

作者： Bo Yang Rui Xie Zhengxun Guo Faculty of Electric Power Engineering Kunming University of Science and TechnologyKunming650500China Intelligent Electric Power Grid Key Laboratory of Sichuan Province Chengdu610065China Foshan Graduate School of Innovation Northeastern UniversityFoshan528311China College of Information Science and Engineering Northeastern UniversityShenyang110819China

Maximum power point tracking(MPPT)technology plays a key role in improving the energy conversion efficiency of photovoltaic(PV)systems,especially when multiple local maximum power points(LMPPs)occur under partial shading conditions(PSC).It is necessary to modify the operating point efficiently and accurately with the help of MPPT technology to maximize the collected *** though a lot of research has been carried out and impressive progress achieved for MPPT technology,it still faces some challenges and ***,the mathematical model established for PV cells is not precise ***,the existing algorithms are often optimized for specific conditions and lack comprehensive adaptability to the actual operating ***,a single algorithm may not be able to give full play to its *** the end,the selection criteria for choosing the suitable MPPT algorithm/converter combination to achieve better performance in a given scenario is very ***,this paper systematically discusses the current research status and challenges faced by PV MPPT technology around the three aspects of MPPT models,algorithms,and hardware *** in-depth thinking and discussion,it also puts forward positive perspectives on future development,and five forward-looking solutions to improve the performance of PV systems MPPT are suggested.

关键词： PV systems MPPT partial shading condition DC-DC converter

来源：评论

学校读者我要写书评

暂无评论

Phenotypic Parameters Extraction of Panax notoginseng with NeRF

Phenotypic Parameters Extraction of Panax notoginseng with N...

引用

information Systems and Computing technology (ISCTech), International Conference on

作者： Ruoxi Wang Liang Dai Qiliang Yang Ling Yang Faculty of Modern Agricultural Engineering Kunming University of Science and Technology Kunming China Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming China

ISBN: (数字)9798350379860

ISBN: (纸本)9798350379877

The growth stage significantly influences both the yield and quality of Panax notoginseng. Accurate plant phenotypic parameters are crucial for the precise management of P. notoginseng cultivation. Currently, traditional methods like manual measurements and tools are used for collecting phenotypic information, but they are inefficient and costly. Manual measurements are prone to subjective biases and can potentially harm plants irreversibly. While 2D images can capture plant phenotypic information, they often suffer from incomplete data and lack precision, especially for plants with complex structures like P. notoginseng. This limitation makes it challenging to achieve comprehensive and accurate measurements. To address these challenges, this study proposes a novel approach using a Neural Radiance Field (NeRF) for extracting phenotypic parameters from P. notoginseng. By capturing video and multi-view image sequences of P. notoginseng, we were able to achieve high-fidelity 3D rendering of the plants and extract point cloud data from them. This approach enabled accurate measurement of plant height and leaf area parameters. The results demonstrate promising accuracy, with an average percentage error of 1.76% for plant height and 1.73% for leaf area based on the point cloud measurements obtained using NeRF. This method leverages advanced computational techniques to overcome the limitations of traditional 2D imaging methods, offering a more comprehensive and precise means of phenotypic characterization for complex plant structures like P. notoginseng.

关键词： Point cloud compression Accuracy Three-dimensional displays Plants (biology) Current measurement Area measurement Crops Manuals Neural radiance field Parameter extraction

来源：评论

学校读者我要写书评

暂无评论

Depth information Assisted Collaborative Mutual Promotion Network for Single Image Dehazing

arXiv

引用

arXiv 2024年

作者： Zhang, Yafei Zhou, Shen Li, Huafeng Faculty of Information Engineering and Automation Kunming University of Science and Technology China

Recovering a clear image from a single hazy image is an open inverse problem. Although significant research progress has been made, most existing methods ignore the effect that downstream tasks play in promoting upstream dehazing. From the perspective of the haze generation mechanism, there is a potential relationship between the depth information of the scene and the hazy image. Based on this, we propose a dual-task collaborative mutual promotion framework to achieve the dehazing of a single image. This framework integrates depth estimation and dehazing by a dual-task interaction mechanism and achieves mutual enhancement of their performance. To realize the joint optimization of the two tasks, an alternative implementation mechanism with the difference perception is developed. On the one hand, the difference perception between the depth maps of the dehazing result and the ideal image is proposed to promote the dehazing network to pay attention to the non-ideal areas of the dehazing. On the other hand, by improving the depth estimation performance in the difficult-to-recover areas of the hazy image, the dehazing network can explicitly use the depth information of the hazy image to assist the clear image recovery. To promote the depth estimation, we propose to use the difference between the dehazed image and the ground truth to guide the depth estimation network to focus on the dehazed unideal areas. It allows dehazing and depth estimation to leverage their strengths in a mutually reinforcing manner. Experimental results show that the proposed method can achieve better performance than that of the state-of-the-art approaches. The source code is released at https://***/zhoushen1/DCMPNet. Copyright © 2024, The Authors. All rights reserved.

关键词： Inverse problems

来源：评论

学校读者我要写书评

暂无评论

A Visual Cryptography Method Based on a New Composite Chaotic System and Latin Square Collaborative Mechanism

SSRN

引用

SSRN 2025年

作者： Tian, Anhong Yue, Zhongze Fu, Chengbiao Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650500 China Faculty of Land Resource Engineering Kunming University of Science and Technology Kunming650093 China

In this study, one-dimensional（1D） chaotic systems are generally characterized by simple chaotic behaviors and discontinuous discrete chaotic intervals. A two-dimensional Composite hyperchaotic System (2D-CHS) construction method is proposed to take advantage of the fact that high-dimensional chaotic systems have more complex dynamics. This innovative method realizes 2D-CHS generation by dimensionally extending multiple 1D seed mappings through a nonlinear coupling mechanism, which provides a more adaptable chaotic system in the field of image encryption. It is demonstrated by the Lyapunov exponent (LE) that the designed 2D-CHS framework is able to exhibit complex hyperchaotic behaviors with robustness. To verify the effectiveness of the framework, three 2D mappings are generated and compared with the existing 2D chaotic systems in multiple metrics, and the results show that the new mappings are significantly improved in performance metrics. Based on the above results, this study further proposes a hybrid cryptographic architecture that fuses chaotic mapping with orthogonal Latin square array. Specifically, the algorithm firstly constructs a Latin square matrix by using short chaotic sequences, and then expands the matrix dimension by Kronecker tensor product operation to generate a high-order orthogonal Latin square matrix matching the size of the target image;finally, it constructs a bi-directional encryption mechanism through the orthogonal characteristics of the matrix, and performs pixel position disruption and grayscale diffusion operations on the image, respectively. The experimental results show that the algorithm not only inherits the advantage of the chaotic system's high sensitivity to the initial value condition, but also gives full play to the unique advantage of the Latin square matrix in orthogonal transformation. In addition, the algorithm has high security and shows important application value in the field of real-time image encryption, providing ne

关键词： Lyapunov functions

来源：评论

学校读者我要写书评

暂无评论

The Tribal Theater Model: Social Regulation for Dynamic User Adaptation in Virtual Interactive Environments

arXiv

引用

arXiv 2024年

作者： Zhang, Hanzhong Duan, Bowen Wang, Haoyang Qiao, Zhijian Yin, Jibin Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming650221 China

This paper proposes a social regulation model for dynamic adaptation according to user characteristics in virtual interactive environments, namely the tribal theater model. The model focuses on organizational regulation and builds an interaction scheme with more resilient user performance by improving the subjectivity of the user. This paper discusses the sociological theoretical basis of this model and how it was migrated to an engineering implementation of a virtual interactive environment. The model defines user interactions within a field that are regulated by a matrix through the allocation of resources. To verify the effectiveness of the tribal theater model, we designed an experimental scene using a chatroom as an example. We trained the matrix as an AI model using a temporal transformer and compared it with an interaction field with different levels of control. The experimental results showed that the tribal theater model can improve users' interactive experience, enhance resilient user performance, and effectively complete environmental interaction tasks under rule-based interaction. © 2024, CC BY-NC-ND.

关键词： Theaters

来源：评论

学校读者我要写书评

暂无评论

An Enhanced Ant Colony Algorithm with Variable Neighborhood Descent for Multi-compartment Vehicle Routing Problem with Time Limits 41

An Enhanced Ant Colony Algorithm with Variable Neighborhood ...

引用

第41届中国控制会议

作者： Ning Guo Bin Qian Jing Na Rong Hu Jian-Lin Mao Academy of Information Engineering and Automation Kunming University of Science and Technology Academy of Mechanical and Electrical Engineering Kunming University of Science and Technology Yunnan Key Laboratory of Artificial Intelligence Kunming University of Science and Technology

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

The multi-compartment vehicle routing problem(MCVRP) is applicable in many areas,such as fuel transportation,garbage collection and food *** colony(ACO) algorithm has been recognized as a prominent evolutionary methodology due to its excellent performance of global *** this paper,an enhanced ant colony algorithm,namely EACO,is presented to minimize the total distance of the MCVRP with time limits(TLs),which is a typical NP-hard combinatorial optimization problem with strong practical application ***,a new pheromone concentration matrix(NPCM) is developed to learn both the information of two successive customers and the customer ***,to improve the quality of initial solutions,three heuristic rules are adopted to initialize the NPCM and the ***,after the ACO-based exploration,a problem-dependent local search with the technique of variable neighborhood descent is presented to perform *** results on benchmark data sets demonstrate the effectiveness of the proposed EACO.

关键词： ant colony algorithm multi-compartment vehicle routing problem time limits variable neighborhood descent

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：