检索结果-内蒙古大学图书馆

remote sensing image Scene Classification Based on Object Relationship Reasoning CNN

IEEE GEOSCIENCE AND remote sensing LETTERS 2022年 19卷 1页

作者： Li, Zhengzhou Wu, Qingqing Cheng, Bei Cao, Lei Yang, Huihui Chongqing Univ Coll Microelect & Commun Engn Chongqing 400044 Peoples R China Chinese Acad Sci Inst Opt & Elect Key Lab Opt Engn Chengdu 610209 Peoples R China

remote sensing image has been widely used in many fields such as military reconnaissance and earthquake relief. However, the complexity and diversity of the scene make the target detection and recognition performance poor. Convolutional neural networks (CNNs) have made breakthrough in remote sensing image processing due to their ability to extract deep features. This letter proposes a remote sensing image scene recognition method based on object relationship reasoning CNN (ORRCNN), which makes use of the relationship between objects to infer the scene information. The method has prior scene-information-based channel and object-detection-based channel to classify the remote sensing image. The prior scene-information-based channel makes use of the feature space to identify the scene, and the object-detection-based channel adopts the relationship between the object and the scene to classify the scene. Afterward, the Bayesian criterion infers the scene more accurately by means of fusing the scene information from the above channels. The experimental results show that the proposed method is excellent especially in the scene where there are iconic objects in the remote image.

关键词： remote sensing Feature extraction image recognition Object detection Cognition Airports Training Convolutional neural network (CNN) object relationship reasoning remote sensing image scene recognition

来源：评论

学校读者我要写书评

暂无评论

Few-shot remote sensing scene classification based on multi subband deep feature fusion

引用

MATHEMATICAL BIOSCIENCES AND ENGINEERING 2023年第7期20卷 12889-12907页

作者： Yang, Song Wang, Huibin Gao, Hongmin Zhang, Lili Hohai Univ Coll Comp & Informat Nanjing 211100 Peoples R China Huaiyin Inst Technol Fac Elect Informat Engn Huaian 223001 Peoples R China

Recently, convolutional neural networks (CNNs) have performed well in object classification and object recognition. However, due to the particularity of geographic data, the labeled samples are seriously insufficient, which limits the practical application of CNN methods in remote sensing (RS) image processing. To address the problem of small sample RS image classification, a discrete wavelet-based multi-level deep feature fusion method is proposed. First, the deep features are extracted from the RS images using pre-trained deep CNNs and discrete wavelet transform (DWT) methods. Next, a modified discriminant correlation analysis (DCA) approach is proposed to distinguish easily confused categories effectively, which is based on the distance coefficient of between-class. The proposed approach can effectively integrate the deep feature information of various frequency bands. Thereby, the proposed method obtains the low-dimensional features with good discrimination, which is demonstrated through experiments on four benchmark datasets. Compared with several state-of-the-art methods, the proposed method achieves outstanding performance under limited training samples, especially one or two training samples per class.

关键词： remote sensing scene classification deep feature fusion discriminant correlation analysis discrete wavelet transform

来源：评论

学校读者我要写书评

暂无评论

Typical Scene Material Classification Based on remote sensing image Data 12

Typical Scene Material Classification Based on Remote Sensin...

引用

SPIE 12th International Symposium on Multispectral image processing and pattern recognition, MIPPR 2023

作者： Gao, Yang Dai, Ling Li, Haipeng Zhang, Yini Wang, Feng Beijing Aerospace Automatic Control Institute Beijing100854 China School of Electrical and Automation Engineering Nanjing Normal University Jiangsu Nanjing210023 China School of Artificial Intelligence and Automation Huazhong University of Science and Technology Hubei Wuhan430074 China

ISBN: (数字)9781510675001

ISBN: (纸本)9781510674998

In spectral scene simulation applications, it is necessary to recognize and detect desert satellite images and satellite images containing water. However, previous network models such as AlexNet, VggNet, and GoogleNet have gradually fallen behind the trend of the times. Compared with the latest network models, they not only have slow training speeds but also do not achieve high accuracy. Therefore, this study adopts a series of the latest network models to solve this problem and compare and analyze the results. © 2024 SPIE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

The Egyptian Journal of remote sensing and Space Sciences

引用

The Egyptian Journal of remote sensing and Space Sciences

作者： Professor Islam Abou El-Magd

The Egyptian Journal of remote sensing and Space Sciences (EJRS) is dedicated to publishing high-quality, peer-reviewed research articles that demonstrate innovative advancements in methodologies, techniques, algorithms, and scientific understanding within the fields of remote sensing and space sciences. EJRS emphasizes contributions that push the boundaries of knowledge and offer transformative solutions to real-world challenges in the following areas: 1. Space and ground remote sensing technology: image and signal processing for: Optical and RADAR image processing theories and technical development image enhancement and restoration, machine learning, data fusion, etc image understanding such as pattern recognition, classification techniques, object and change detection, and content-based retrieval of images Geophysical measurements and LIDAR image processing theories and technical development 2. Advanced Applications of remote sensing and GIS in: Geology: New methodologies for planetary geology, geomorphology, structural geology, hydrogeology, mineral exploration, space archaeology, and coastal/marine geology Agriculture: Soil Science - Crop conditions assessment, Development of precision agriculture techniques, innovative irrigation management systems, advanced tools for crop yield estimation and agro-environmental protection Urban Planning: Novel GIS-based solutions for highway layouts, site investigations, engineering soil mapping, and urban corridor surveys Hydrology: Advanced modeling techniques for water resource management - dam sites location - flash flood analysis - water quality modeling and monitoring Environment: Land Degradation assessment- Disasters - Desertification-- Coastal and Marine Environment - climate change impact studies - Land Evaluation Geomatics: Innovations in digital photogrammetry, LIDAR, hyperspectral imaging, change detection, and feature extraction Machine Learning and AI: Pioneering applications in natural hazard assessment, mitig

关键词：

来源：评论

学校读者我要写书评

暂无评论

Effective processing of Convolutional Neural Networks for Computer Vision: A Tutorial and Survey

引用

IETE TECHNICAL REVIEW 2022年第1期39卷 49-62页

作者： Tombe, Ronald Viriri, Serestina Univ KwaZulu Natal Comp Sci Discipline Durban South Africa

Over the past few years, Convolutional neural networks (ConvNets) is emerging as computer vision discipline within deep learning. ConvNets is a key strategy for addressing computer vision problems, yet the theories behind their effectiveness in the processing are not yet fully understood. ConvNets have attained a state of the art-performances on various datasets for computer vision tasks such as remote sensing images scene classification, face recognition, and object detection. This is attributed to their effectiveness in image feature processing. This work reviews the major advances on ConvNets for effective processing in computer vision from some dimensions which include, convolutional layer design configurations, pooling layer strategies, network activation functions, loss functions, network regularization techniques, and ConvNet optimization methods. Further, this works surveys the application of ConvNets on three computer vision tasks,*** sensing images scene recognition, face recognition, and object detection to demonstrate the effectiveness of convNets in image feature processing. Additionally, datasets for evaluation and benchmarking purposes with the aforementioned computer vision tasks are briefly discussed.

关键词： Convolutional neural networks Face recognition Object detection remote sensing

来源：评论

学校读者我要写书评

暂无评论

Adaptive Spatial Complex Fuzzy Inference Systems With Complex Fuzzy Measures

引用

IEEE ACCESS 2023年 11卷 39333-39350页

作者： Giang, Le Truong Son, Le Hoang Giang, Nguyen Long Luong, Nguyen Van Lan, Luong Thi Hong Tuan, Tran Manh Thang, Nguyen Truong Hanoi Univ Ind Ctr Qual Assurance Hanoi 100000 Vietnam Vietnam Acad Sci & Technol Grad Univ Sci & Technol Hanoi 100000 Vietnam Vietnam Natl Univ VNU VNU Informat Technol Inst Hanoi 700000 Vietnam Vietnam Acad Sci & Technol Inst Informat Technol IoIT Hanoi 100000 Vietnam Thuyloi Univ Fac Comp Sci & Engn Hanoi 116705 Vietnam

Fuzzy inference systems, in general, and complex fuzzy inference systems, in particular, play an increasingly important role in many fields, such as change detection, image classification, recognition problems, etc. Despite being the well-known technique to solve with time series data, the rulebase still has the considered limitation because of the directly affecting the results as well as the processing time of these methods. To overcome this limitation, this study proposes an Adaptive spatial complex inference system that can automatically infer and adapt to the new remotely sensed image. In the proposed model, to predict the image of time t + 1, the system will generate a new rulebase according to this expected image. This new rulebase and the previous Co-Spatial-CFIS+ rulebase are evaluated using a complex fuzzy measure. This measure is built by determining the intersection domain between two rule spaces;this intersection value estimates removing, merging, or adding a newly generated rule into the current rulebase. Finally, a more suitable set of rules is obtained for image prediction. To illustrate the efficiency of the proposed approach, it is applied to the remote sensing cloud image data of the U.S. Navy. Our model evaluated the model's effectiveness in comparison to the state-of-the-art along studies in detecting changes in remote sensing cloud images. Moreover, the findings of the experiments revealed that the proposed model could improve the change detection results in terms of R-2, RMSE, time-consuming, and the number of rules.

关键词： Fuzzy logic Adaptation models remote sensing Adaptive systems Time series analysis Predictive models Fuzzy sets Complex fuzzy inference system remote sensing images rule pruning rule-based system image change detection

来源：评论

学校读者我要写书评

暂无评论

Algorithmic Research on Hydrogeological Comprehensive Monitoring Using Multi-Source remote sensing Data 24

Algorithmic Research on Hydrogeological Comprehensive Monito...

引用

International Conference on Machine Intelligence and Digital Applications (MIDA)

作者： Yuan, Yuan Shandong Prov Bur Geol & Mineral Resources Jinan 250014 Shandong Peoples R China Shandong Engn Res Ctr Environm Protect & Remediat Jinan 250014 Shandong Peoples R China

ISBN: (纸本)9798400718144

In response to the difficulties in obtaining data and high-error data in current hydrogeological surveys, this article uses high-resolution satellite image processing to capture wastewater characteristics based on the principle of remote sensing reflectance. The automatic threshold selection algorithm determines the optimal threshold for extracting wastewater information and establishes a model for identifying wastewater, achieving the purpose of better identification of water quality. This article uses drones to capture geological images for processing. Based on two-dimensional wavelet transform, remote sensing images are divided into four sub bands and filtered. The four sub bands can be input into the residual network learning algorithm to obtain high-quality four new sub bands, and finally transformed into a complete high-quality remote sensing geological image through inverse two-dimensional wavelet transform, achieving the purpose of better identification of water quality. The results show that the wastewater recognition model can accurately identify wastewater according to the characteristics that the reflectance of wastewater water in blue. Green and red wavelengths is lower than that of normal water, while the reflectance in infrared wavelengths is higher than that of normal water. The peak signal-to-noise ratio and mean square error of the residual learning network algorithm using two-dimensional wavelet transform for geological remote sensing images are 28.12 and 0.051, which are the best compared with the other two methods, indicating that the image quality after processing is high.

关键词： Hydrogeological Study Multi-Source remote sensing Automatic Threshold Selection Two-dimensional Wavelet Transform Residual Network Learning

来源：评论

学校读者我要写书评

暂无评论

DICMNet: Dynamic Irregular Resnet with Multi-direction Channel Remapping for remote sensing Road Extraction 7th

DICMNet: Dynamic Irregular Resnet with Multi-direction Chann...

引用

7th Chinese Conference on pattern recognition and Computer Vision

作者： Li, Bowen Xiao, Feng Liu, Siyu Shen, Chao Xian Technol Univ Sch Comp Sci & Engn Xian 710021 Peoples R China Xian Technol Univ Sch Weap Sci & Technol Xian 710021 Peoples R China

ISBN: (纸本)9789819784929;9789819784936

Road information construction is the key support to solve urban traffic problems. The technique of segmenting the road part from the image is called road extraction. Therefore, road extraction can be used as a technical support for traffic system to obtain a large amount of road information. Due to the advantages of high resolution and overhead shooting angle for remote sensing image, it is very suitable as the data source of road extraction. Although road extraction from remote sensing images becomes one of the important research directions and some achievements have been made, there are still many challenges. Different from common objects, roads have strong topological continuity and are also composed of many irregular curves. This special morphological feature is difficult to be mined by common convolution kernel. To solve the above problems, this paper proposes a novel feature extraction backbone network DICMNet, which is composed of dynamic irregular convolutional (DIC) module and multi-directional channel remapping (MDCR) combine with resnet. Experimental validation with the other methods is performed on the DeepGlobe road dataset. Our model has a significant improvement on most of the evaluation metrics with an average improvement of 2%.

关键词： remote sensing road extraction dynamic irregular convolution multi-direction channel remapping

来源：评论

学校读者我要写书评

暂无评论

Partial Discharge pattern recognition Method Based on Transfer Learning and DenseNet Model

引用

IEEE TRANSACTIONS ON DIELECTRICS AND ELECTRICAL INSULATION 2023年第3期30卷 1240-1246页

作者： Fu, Yuwei Liang, Liejuan Huang, Weihua Huang, Guobin Huang, Peijun Zhang, Zhiyu Chen, Chi Wang, Chuang Xian Univ Technol Sch Elect Engn Xian 710018 Peoples R China Jiangmen Guobin Robot Co Ltd Jiangmen 529099 Peoples R China Jiangmen Reignwood Technol Co Ltd Jiangmen 529099 Peoples R China

With the development of intelligent sensing technology, a large amount of partial discharge (PD) time-domain waveform images are generated in the on-site detection of gas-insulated switchgear (GIS) PD. Traditional pattern recognition methods are mostly aimed at structured data and cannot directly identify defect types of such data. At the same time, the deep learning method for GIS PD pattern recognition is generally faced with the problem of small samples. In order to solve the above problems, this article proposes a PD pattern recognition method based on transfer learning and DenseNet model. First, the time-domain waveform images are processed by image enhancement, normalization, image compression, and other image processing techniques. The finite-difference time-domain (FDTD) method was used to simulate GIS PD, and the time-domain waveform image database of four PD defects is established. Using convolutional neural network (CNN) and transfer learning, the recognition accuracy of the model is increased to 95%, with better robustness. The recognition performance of different CNN structures is studied. The results show that DenseNet model has higher accuracy than other structures and shorter training time. This study can be used to diagnose the insulation status of GIS equipment in-site.

关键词： Data models pattern recognition Feature extraction Transfer learning Time-domain analysis Partial discharges Gas insulation DenseNet gas-insulated switchgear (GIS) partial discharge (PD) pattern recognition transfer learning

来源：评论

学校读者我要写书评

暂无评论

Vertical Synchrosqueezing for High-Resolution Radar Imaging

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND remote sensing 2023年 61卷 1-13页

作者： Abratkiewicz, Karol Drozdowicz, Jedrzej Samczynski, Piotr Warsaw Univ Technol Inst Elect Syst Fac Elect & Informat Technol PL-00665 Warsaw Poland

This article presents a novel approach to inverse synthetic aperture radar (ISAR) image enhancement using time-frequency (TF) vertical synchrosqueezing (VSS). The authors proposed adding the signal extraction block to the ISAR image processing path. Since the VSS technique does not change the signal amplitude-phase dependencies but only performs its decomposition and denoising, the proposed method is noninvasive. It was shown that denoising the signal by the VSS in the TF domain reduces the sidelobe level of the ISAR image and enables its denoising. For imaging purposes, the authors derived four novel TF signal parameter estimators in the TF domain that are practically investigated in a radar application. The proposed technique was extensively tested through simulation studies and on data obtained from measurements using high-resolution wideband frequency-modulated continuous wave (FMCW) radar. The usefulness of the proposed technique is shown using simulated and real-life results.

关键词： Inverse synthetic aperture radar (ISAR) radar remote sensing time-frequency (TF) signal analysis

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：