检索结果-内蒙古大学图书馆

7th IEEE International Conference on Information Systems and Computer Aided Education, ICISCAE 2024

作者： Xu, Fan College of Computer Science and Technology Ocean University of China Qingdao266404 China

ISBN: (纸本)9798350350760

With the widespread application of remote sensing technology, accurate and rapid detection and identification of targets in remote sensing images have become an important research area. Traditional target detection methods often face issues with low recognition accuracy and slow response when dealing with complex backgrounds or small targets. This study proposes an improved YOLO (You Only Look Once) model integrated with a dual attention mechanism to enhance the performance of target detection in remote sensing images. This mechanism utilizes spatial attention to increase the model's sensitivity to target locations, while channel attention enhances the recognition of target features, enabling more accurate identification of small and complex environment targets. The model was tested across multiple remote sensing image datasets, and the results show significant improvements in detection accuracy and real-time performance compared to traditional YOLO models. Additionally, the implementation of the model reveals the potential of the dual attention mechanism in improving the efficiency and accuracy of remote sensing image processing. With further optimization and adjustment, this model is expected to play a significant role in various application scenarios such as environmental monitoring, urban planning, and disaster assessment. ©2024 IEEE.

关键词： Environmental monitoring

来源：评论

学校读者我要写书评

暂无评论

Spatial-spectral unfolding network with mutual guidance for multispectral and hyperspectral image fusion

引用

pattern recognition 2025年 161卷

作者： Yan, Jun Zhang, Kai Sun, Qinzhu Ge, Chiru Wan, Wenbo Sun, Jiande Zhang, Huaxiang Shandong Normal Univ Sch Informat Sci & Engn Jinan Peoples R China Xidian Univ Sch MicroElect Xian Peoples R China

Fusing low spatial resolution hyperspectral (LR HS) and high spatial resolution multispectral (HR MS) images from different modalities aim to obtain high spatial resolution hyperspectral (HR HS) images. However, most deep neural network (DNN)-based methods overlook the correlation between the spatial domain and spectral domain, leading to limited fusion performance. To solve this problem, we propose the spatial-spectral unfolding network with mutual guidance (SMGU-Net). Specifically, the information of different modalities in the source images is treated as mutual complementary components to derive the reconstruction model. Then, the model is optimized using half-quadratic splitting and gradient descent algorithms and is unfolded into a network that leverages the powerful learning capabilities of DNNs to explore more potential information in the deep feature space. In this way, the network achieves the interaction and supplementarity of cross-modality information generate fused images. Experiments are conducted on four benchmark datasets to demonstrate the effectiveness of SMGU-Net. The code can be downloaded from https://***/yansql/SMGU-Net.

关键词： remote sensing Unfolding network image fusion Multispectral image Hyperspectral image

来源：评论

学校读者我要写书评

暂无评论

WildFishNet: Open Set Wild Fish recognition Deep Neural Network With Fusion Activation pattern

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND remote sensing 2023年 16卷 7303-7314页

作者： Zhang, Xiaoya Huang, Baoxiang Chen, Ge Radenkovic, Milena Hou, Guojia Qingdao Univ Dept Comp Sci & Technol Qingdao 266071 Peoples R China Qingdao Natl Lab Marine Sci & Technol Lab Reg Oceanog & Numer Modeling Qingdao 266228 Peoples R China Qingdao Natl Lab Marine Sci & Technol Lab Reg Oceanog & Numer Modeling Qingdao 266228 Peoples R China Ocean Univ China Sch Marine Technol Inst Adv Ocean Study Qingdao 266075 Peoples R China Univ Nottingham Sch Comp Sci & Informat Technol Nottingham NG8 1BB England

Wild fish recognition is a fundamental problem of ocean ecology research and contributes to the understanding of biodiversity. Given the huge number of wild fish species and unrecognized category, the essence of the problem is an open set fine-grained recognition. Moreover, the unrestricted marine environment makes the problem even more challenging. Deep learning has been demonstrated as a powerful paradigm in image classification tasks. In this article, the wild fish recognition deep neural network (termed WildFishNet) is proposed. Specifically, an open set fine-grained recognition neural network with a fused activation pattern is constructed to implement wild fish recognition. First, three different reciprocal inverted residual structural modules are combined by neural structure search to obtain the best feature extraction performance for fine-grained recognition;next, a new fusion activation pattern of softmax and openmax functions is designed to improve the recognition ability of open set. Then, the experiments are implemented on the WildFish dataset that consists of 54 459 unconstrained images, which includes 685 known classes and 1 open set unrecognized category. Finally, the experimental results are analyzed comprehensively to demonstrate the effectiveness of the proposed method. The in-depth study also shows that artificial intelligence can empower marine ecosystem research.

关键词： Deep neural network fusion activation pattern neural structure search open set fine-grained recognition wild fish recognition

来源：评论

学校读者我要写书评

暂无评论

Self-supervised polarization image dehazing method via frequency domain generative adversarial networks

引用

pattern recognition 2025年 165卷

作者： Sun, Rui Chen, Long Liao, Tanbin Fan, Zhiguo Hefei Univ Technol Sch Comp Sci & Informat Engn 485 Danxia Rd Hefei 230009 Peoples R China Hefei Univ Technol Key Lab Ind Safety & Emergency Technol Hefei 230009 Peoples R China Minist Educ Peoples Republ China Key Lab Knowledge Engn Big Data Hefei 230009 Peoples R China

Haze significantly hinders the application of autonomous driving, traffic surveillance, and remote sensing. image dehazing serves as a key technology to enhance the clarity of images captured in hazy conditions. However, the lack of paired annotated training data significantly limits the performance of deep learning-based dehazing methods in real-world scenarios. In this work, we propose a self-supervised polarization image dehazing framework based on frequency domain generative adversarial networks. By incorporating a polarization calculation module into the generator, the Stokes parameters of airlight are accurately estimated, which are used to reconstruct the synthesized hazy image by combining the dehazed image generated via a densely connected encoder-decoder. Furthermore, we optimize the discriminator with frequency domain features extracted by frequency decomposition module and introduce a pseudo airlight coefficient supervision loss to enhance the selfsupervised training. By discriminating between synthetic hazy images and real hazy images, we achieve adversarial training without the need for paired data. Simultaneously, supervised by the atmospheric scattering model, our network can iteratively generate more realistic dehazed images. Extensive experiments conducted on the constructed multi-view polarization datasets demonstrate that our method achieves state-of-the-art performance without requiring real-world ground truth.

关键词： image dehazing Polarization image Frequency domain Self-supervised Generative adversarial network

来源：评论

学校读者我要写书评

暂无评论

Segment Anything Model for Road Network Graph Extraction

Segment Anything Model for Road Network Graph Extraction

引用

IEEE/CVF Conference on Computer Vision and pattern recognition (CVPR)

作者： Hetang, Congrui Xue, Haoru Le, Cindy Yue, Tianwei Wang, Wenping He, Yihui Carnegie Mellon Univ Pittsburgh PA 15213 USA Columbia Univ New York NY USA

ISBN: (纸本)9798350365474

We propose SAM-Road, an adaptation of the Segment Anything Model (SAM) [27] for extracting large-scale, vectorized road network graphs from satellite imagery. To predict graph geometry, we formulate it as a dense semantic segmentation task, leveraging the inherent strengths of SAM. The image encoder of SAM is fine-tuned to produce probability masks for roads and intersections, from which the graph vertices are extracted via simple non-maximum suppression. To predict graph topology, we designed a lightweight transformer-based graph neural network, which leverages the SAM image embeddings to estimate the edge existence probabilities between vertices. Our approach directly predicts the graph vertices and edges for large regions without expensive and complex post-processing heuristics and is capable of building complete road network graphs spanning multiple square kilometers in a matter of seconds. With its simple, straightforward, and minimalist design, SAM-Road achieves comparable accuracy with the state-of-the-art method RNGDet++[57], while being 40 times faster on the City-scale dataset. We thus demonstrate the power of a foundational vision model when applied to a graph learning task. The code is available at https://***/htcr/sam_road.

关键词： autonomous driving computer vision foundation model graph graph neural network mapping navigation remote sensing segment anything semantic segmentation transformer

来源：评论

学校读者我要写书评

暂无评论

Development of Broad Area Target Search System Based on Deep Learning 2

Development of Broad Area Target Search System Based on Deep...

引用

2023 2nd International Conference on Environmental remote sensing and Geographic Information Technology, ERSGIT 2023

作者： Gao, Yan Lu, Donghua Zhang, Yiting Wang, Wei National Key Laboratory of Remote Sensing Information and Imagery Analyzing Technology Beijing Research Institute of Uranium Geology Beijing100029 China

ISBN: (纸本)9781510672949

Deep learning is increasingly being applied in the field of remote sensing image processing. However, researchers often face limitations in establishing high-quality datasets for target recognition in high-resolution imagery due to resource constraints. Moreover, traditional target recognition methods cannot leverage the extensive coverage of remote sensing images for broad-scope searches of specific targets. To address the scarcity of high-resolution remote sensing image sources for target detection tasks, achieve broad-scope automatic searching for areas of interest, and enhance the level of task automation, a system with the capability to acquire data from multiple network sources and recognize nuclear power plant targets was developed. The system initially integrates data from multiple network map servers, enabling the rapid and cost-effective acquisition of high-resolution, large-area remote sensing imagery. Deep learning samples are created based on multi-source image data, and multiple target detection models are trained. The software's user interface was developed and tested, facilitating broad-scope searches for key components of nuclear power stations based on extensive network imagery. © 2024 SPIE.

关键词： Nuclear power plants

来源：评论

学校读者我要写书评

暂无评论

A Vision-based remote Assistance Method and it's Application in Object Transfer 2

A Vision-based Remote Assistance Method and it's Application...

引用

2nd Asia Conference on Computer Vision, image processing and pattern recognition (CVIPPR)

作者： Cheng, Mingkai Yi, Pengfei Guo, Yujie Liu, Rui Dong, Jing Zhou, Dongsheng Dalian Univ Sch Software Engn Key Lab Adv Design & Intelligent Comp Minist Educ Dalian Liao Ning Peoples R China Dalian Univ Technol Sch Comp Sci & Technol Dalian Liao Ning Peoples R China

ISBN: (纸本)9798400716607

remote assistance by users is a common means in current robot applications, which can improve the efficiency of transfer tasks. However, the current mainstream methods are costly, relying on auxiliary equipment, and result in a significant workload for users throughout the task. To reduce the cost and workload of remote assistance and improve user interaction with robots. This paper has undertaken the following work using available visual image information. Firstly, proposing a remote guidance method based on facial information, which is derived from analyzing user facial information for gaze estimation and combined with image detection. Secondly, proposing an autonomous transfer method based on position inference, which combines target information in images with the real-time state of the robot. Finally, by integrating the above methods, proposing a vision-based remote assistance method and applyingit in real-world experiments of object transfer tasks.

关键词： Visual processing Gaze Estimation Position Inference Object Transfer remote Assistance

来源：评论

学校读者我要写书评

暂无评论

Research on remote sensing image processing Model Based on Convolutional Neural Net 6

Research on Remote Sensing Image Processing Model Based on C...

引用

6th International Conference on Information Technologies and Electrical Engineering, ICITEE 2023

作者： Wang, Zhe School of Information Engineering Wuhan University of Technology Wuhan China

ISBN: (纸本)9798400708299

With the ongoing advancements in artificial intelligence technology, Convolutional Neural Networks (CNNs) have found widespread application in the realm of remote sensing image processing. They prove instrumental in tasks such as feature categorization, target detection, and change detection. Despite these strides, the continuous evolution of deep learning-based technology for remote sensing image processing confronts intricate and specific challenges, including the scarcity of high-quality remote sensing image data, strategies for handling multimodal data, and the detection of small targets in images. This paper proposes a remote sensing image processing model grounded in Convolutional Neural Networks to enhance the classification accuracy of remote sensing images. The approach incorporates the use of bilateral filtering to address interference factors in the image, preserving crucial image details. Furthermore, point operation-based histogram equalization is employed for image processing, facilitating the extraction of inconspicuous features within the image. The proposed model undergoes testing, and after more than 40 training iterations, it achieves a classification accuracy of over 90% for remote sensing images. The image processing results demonstrate effectiveness, showcasing high model recognition accuracy. © 2023 ACM.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

remote sensing image change detection method based on hybrid trunk and high and low frequency attention 4

Remote sensing image change detection method based on hybrid...

引用

4th International Conference on image processing and Intelligent Control, IPIC 2024

作者： Jiang, Qiuyue State Key Laboratory on Integrated Optoelectronics College of Electronic Science & Engineering Jilin University Changchun130012 China

ISBN: (纸本)9781510682313

remote sensing image change detection technology is rapidly advancing under the impetus of deep *** this study, a remote sensing image change detection method based on hybrid backbone and high and low frequency attention modules is proposed for the problems of insufficient extraction of feature components of traditional convolutional neural networks in this field of remote sensing image change *** method adopts a hybrid trunk network, and through the attention feature fusion module, the features of the two branches of convolutional neural network and Transformer are fused, which can take care of the extraction of local features and global information. Further, this study integrates high and low frequency attention modules to refine the high frequency details and low frequency background information in the image respectively. The implementation of this method significantly improves the quality and depth of feature extraction. Ultimately, the ability to discriminate and extract the location information and features of interest is strengthened by the coordinate attention module, which improves the recognition accuracy of local *** extensive experimental testing and validation, it is confirmed that the proposed model achieves a significant improvement in performance compared to existing change detection models. © 2024 SPIE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Third International Conference on Computer Vision and pattern Analysis, ICCPA 2023

Third International Conference on Computer Vision and Patter...

引用

3rd International Conference on Computer Vision and pattern Analysis, ICCPA 2023

ISBN: (纸本)9781510667563

The proceedings contain 140 papers. The topics discussed include: digital multi-scale visual planning model of spatial-geographical landscape pattern of smart parks;visual question answering model based on fusing global-local feature;image processing of the special sensor microwave/imager based on passive microwave remote sensing;image processing of the special sensor microwave/imager based on passive microwave remote sensing;graptolite image classification based on feature transfer and mixup data enhancement;an image classification method based on few-shot learning;fine-grained image recognition based on multi-branch and multi-scale learning;research on road extraction model of remote sensing image based on the fused convolutional module and attention mechanism;unsupervised aircraft detection in SAR images with image-level domain adaption from optical images;the role of echocardiography segmentation evaluation metrics in clinical diagnosis;and machine vision-based measurement of air compressor crankshaft journal dimensions.

关键词：

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：