检索结果-内蒙古大学图书馆

IET International Radar Conference 2023, IRC 2023

作者： Wang, Zhibin Zhang, Qingjun Lv, Zheng Liu, Jiuli Zhou, Yashi Yang, Taoli Institute of Remote Sensing Satellite China Academy of Space Technology Beijing China University of Electronic Science and Technology of China Chengdu China

ISBN: (纸本)9781839539954

Multi-angle Synthetic Aperture Radar (SAR) is a new SAR working mode to acquire multi-dimensional information of target. By observing the target area from different observation angles, the multi-angle SAR has the ability to acquire moving target information. This paper constructs a multi-angle SAR moving target observation signal model, derives a method for estimating the speed of a moving target based on the SAR image acquired from forward and backward observation, and obtains the target azimuthal and radial speeds using this method. The validity of this method is verified through simulation 濸瀋瀃濸瀅濼瀀濸瀁瀇, and it can be used to accurately estimate the speed of moving targets. © The Institution of Engineering & Technology 2023.

关键词： Radar target recognition

来源：评论

学校读者我要写书评

暂无评论

Terrain image Classification Based on Vision Transformer Deep Learning Algorithm 7

Terrain Image Classification Based on Vision Transformer Dee...

引用

7th International Conference on Algorithms, Computing and Artificial Intelligence, ACAI 2024

作者： Ren, Shiqi Zhang, Yu College of information engineering College of artificial intelligence Yangzhou university Jiangsu Yangzhou225000 China College of Computing Georgia Institute of Technology AtlantaGA30332 United States

ISBN: (纸本)9798331529314

Terrain image classification is an important research direction in the field of remote sensing and computer vision, aiming to realize automatic recognition and classification of different landform features through the analysis and processing of terrain images. In this paper, a deep learning algorithm based on Vision Transformer (ViT) is used to classify terrain images, and the performance of the algorithm in this task is systematically evaluated. In the process of model construction, we first imported the Vision Transformer model and made corresponding parameter Settings to ensure its adaptability and effectiveness. After training, it is observed that the loss function of the training set decreases from the initial value of 2.84 to 0.35, a decrease of 2.49, indicating that the model tends to converge in continuous optimization. At the same time, the accuracy is also significantly improved, from 55.9% to 86.8%, an increase of 30.9%, showing the enhancement of the model's learning ability. For the validation set, loss also decreased from 0.78 to 0.47, a decrease of 0.31, while accuracy increased from 60.1% to 83.5%, an increase of 23.4%. These results further prove the good performance of the model on different data sets and its convergence trend. In addition, through the evaluation of the test set, we get more specific performance indicators: The accuracy of the terrain image classification model based on Vision Transformer on the test set reaches 89.9%, the Precision is 0.9615, the Recall is 0.9494, and the F1-score is 0.9554. These indicators show that the model not only has high classification accuracy, but also performs well in generalization ability. To sum up, this research demonstrates the effectiveness of Vision Transformer deep learning algorithm in terrain image classification, and provides a new solution idea and method for related fields. Through continuous optimization and adjustment, the algorithm is expected to achieve more extensive promotion in pract

关键词： Machine vision

来源：评论

学校读者我要写书评

暂无评论

Multi-View Stereo and Depth Priors Guided NeRF for View Synthesis 26

Multi-View Stereo and Depth Priors Guided NeRF for View Synt...

引用

26th International Conference on pattern recognition / 8th International Workshop on image Mining - Theory and Applications (IMTA)

作者： Deng, Wang Zhang, Xuetao Guo, Yu Lu, Zheng Xi An Jiao Tong Univ Inst Artificial Intelligence & Robot Xian Peoples R China Xi An Jiao Tong Univ Sch Software Engn Xian Peoples R China China Acad Space Technol Inst Remote Sensing Satellite Xian Peoples R China

ISBN: (数字)9781665490627

ISBN: (纸本)9781665490627

In this paper, we present a new framework for view synthesis of novel view based on Neural Radiance Fields(NeRF). We aim to address two main limitations of NeRF. Firstly, we propose to combine multi-view stereo into NeRF to help construct general neural radiance fields across different scenes. Specifically, We build a MVS-Encoding Feature Volume with average group-wise correlation to aggregate the multi-view appearance and geometry feature for every source view. And then we use an MLP to encode neural radiance fields by using the scene-dependent features interpolated from the MVS-Encoding Feature Volumes. This makes our model can be applied to other unseen scenes without any per-scene fine-tuning, and render realistic images with few images. If more training images are provided, our method can be fine-tuned quickly to render more realistic images. In fine-tuning phase, we propose a depth priors guided sampling method, which can make the model represent more accurate geometry for corresponding scenes and so render high-quality images of novel view. We evaluate our method on three common datasets. The experiment results show that our method performs better than other baselines, neither without or with fine-tuning. And the depth priors guided sampling method can be easily applied on other methods based on Neural Radiance Fields to further improve the quality of rendered images.

关键词： View Synthesis Neural Radiance Fields Multi-View Stereo Depth Priors

来源：评论

学校读者我要写书评

暂无评论

Towards Tackling Multi-Label Imbalances in remote sensing imagery 25

Towards Tackling Multi-Label Imbalances in Remote Sensing Im...

引用

25th International Conference on pattern recognition (ICPR)

作者： Kossmann, Dominik Wilhelm, Thorsten Fink, Gernot A. TU Dortmund Univ Dept Comp Sci Dortmund Germany TU Dortmund Univ Dept Elect Engn Dortmund Germany

ISBN: (纸本)9781728188089

Recent advances in automated image analysis have lead to an increased number of proposed datasets in remote sensing applications. This permits the successful employment of data hungry state-of-the-art deep neural networks. However, the Earth is not covered equally by semantically meaningful classes. Thus, many land cover datasets suffer from a severe class imbalance. We show that by taking appropriate measures, the performance in the minority classes can be improved by up to 20 percent without affecting the performance in the majority classes strongly. Additionally, we investigate the use of an attribute encoding scheme to represent the inherent class hierarchies commonly observed in land cover analysis.

关键词： Applications of computer vision Multilabel learning Scene understanding

来源：评论

学校读者我要写书评

暂无评论

remote-sensing Based Precipitation Detection Using Conditional GAN and Recurrent Neural Networks 27th

Remote-Sensing Based Precipitation Detection Using Condition...

引用

27th Iberoamerican Congress on Progress in pattern recognition, image Analysis, Computer Vision, and Applications, CIARP 2024

作者： Negri, Pablo Silvarrey, Alejo Gonzalez, Sergio Ruiz, Juan Vidal, Luciano UBA-CONICET Buenos Aires Argentina Departamento de Computación FCEyN UBA Buenos Aires Argentina Universidad Católica del Uruguay Punta del Este Uruguay Departamento de Ciencias de la Atmósfera y los Océanos FCEyN UBA Buenos Aires Argentina UBA-CONICET Buenos Aires Argentina Buenos Aires Argentina Buenos Aires Argentina

ISBN: (纸本)9783031766039

Precipitation detection using infrared (IR) brightness temperature (BT) temporal flux data is a challenging problem. Other sensors, such as microwave (MW), have reliable and more robust predictive performance, but lack land coverage and temporal availability. IR-BT provides high-frequency data (from half an hour to 10 min) at very low resolution (4 km). However, automatic precipitation detection frameworks should face the simple nature of this variable on the one hand, and the very low number of rain events occurring in nature on the other hand. This paper addresses this challenge by proposing a conditional GAN framework using recurrent neural networks, which transforms the unbalanced problem into a small (short) pattern detection algorithm. Several tests allow the identification of robust architectures and useful loss functions that enable promising results, minimize false alarms, and improve the overlap of positive events. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

SDNet: Sandwich decoder network for waterbody segmentation in remote sensing imagery

引用

IEEE Journal of Selected Topics in Applied Earth Observations and remote sensing 2025年 18卷 13895-13911页

作者： Ni, Hao Li, Jianfeng Wang, Chenxu Zhou, Zhiquan Wang, Xinsheng Harbin Institute of Technology at Weihai School of Information Science and Engineering Weihai264209 China

Waterbody extraction is essential for monitoring surface changes and supporting disaster response. However, differences in morphology, dimensions, and spectral reflectance, make it problematic to segregate waterbodies accurately in remote sensing (RS) photographs. Although self-attention (SA) and convolutional neural networks (CNN) have demonstrated strong modeling capabilities on RS image recognition, their efficiency and expressiveness still deserve further optimization. In this paper, we propose SDNet, an innovative deep learning structure designed to improve segmentation accuracy and scalability in waterbody extraction while minimizing computational costs. SDNet introduces a Two-Step Self-Attention (TSSA) strategy to selectively aggregate spatial features and a Sparse-Tight pattern Fusion (STPF) module to capture context across various manipulation spaces, enhancing axial and space information fusion. The network's sandwich-shaped decoder structure allows for easy integration with mainstream encoders. Compared to previous influential methods, SDNet exhibits superior sensitivity in detecting small waterbodies and delivers better overall performance. Experiments performed on public benchmark datasets, like MSRWD, UAVid, and LoveDA, indicate that SDNet exceeds prior state-of-the-art techniques, with a 2%-10% increase in mIoU and a 3%-9% improvement in F1-score. On the MSRWD, SDNet attained an IoU of 94.21% for the water class, with an F1-score of 97.03% and an overall accuracy of 98.87%. © 2008-2012 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

HOW SAR image DENOISE AFFECTS THE PERFORMANCE OF DCNN-BASED TARGET recognition METHOD

HOW SAR IMAGE DENOISE AFFECTS THE PERFORMANCE OF DCNN-BASED ...

引用

IEEE International Geoscience and remote sensing Symposium (IGARSS)

作者： Tang, Jiaxin Zhang, Fan Ma, Fei Gao, Fei Yin, Qiang Zhou, Yongsheng Beijing Univ Chem Technol Coll Informat Sci & Technol Beijing 100029 Peoples R China

ISBN: (纸本)9781665403696

Currently, deep neural networks have been widely used in the field of SAR target recognition. Many researchers found that deep neural networks have an ability of denoising. In many cases, there is no need to denoise in pre-process. But the denoising ability of deep neural networks can take place of conventional denoising algorithm or not is doubtful. In this article, we explore the effect of image denoising algorithms to SAR target recognition methods based on deep neural networks. Firstly, seven traditional denoising algorithms are selected to process two SAR datasets. And these data are utilized to train two kinds of deep neural networks. After comparing and analyzing the training processes and results, we find that 1) The effect of denoising algorithms is influenced by architectures of neural networks and quality of datasets. It is difficult to find a SAR image denoising algorithm, which can improve the accuracy of any recognition network. Sometimes they even drag down the performance of recognition networks. 2) The deep networks with more layers will have better denoising ability, so the effect of denoising algorithms will decrease. For ResNet, there is no need to add the denoising processing.

关键词： Synthetic aperture radar (SAR) speckle denoising algorithm deep neural networks target recognition

来源：评论

学校读者我要写书评

暂无评论

Object recognition of Urban Buildings in High Spatial Resolution remote sensing imagery 2

Object Recognition of Urban Buildings in High Spatial Resolu...

引用

2021 2nd International Conference on Geology, Resources Exploration and Development, ICGRED 2021

作者： Liu, Yuan School of Geomatics and Urban Spatial Information Beijing University of Civil Engineering and Architecture Beijing100044 China

High-resolution remote sensing images can finely express rich surface information. Using the macroscopic and spatial-temporal full coverage advantages of high-resolution remote sensing images for urban building objectification recognition is a current research hot spot in the field of remote sensing image analysis and application. However, the current research still lacks effective technical means to convert surface elements quickly and accurately from remote sensing image space to geographic information space. In this paper, the complementary advantages between image processing and deep learning are combined to research target-level extraction of high-resolution remote sensing urban buildings based on the building element information data. The experimental results achieved the precision of 0.9481 and the recall of 0.9733, indicating that the method proposed in this paper can be applied to the effective extraction of urban-level buildings, which expands the theoretical basis of the object recognition method of remote sensing thematic features based on the idea of object image analysis. © Published under licence by IOP Publishing Ltd.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

remote sensing image Registration Via Feature Fusion and Enhanced Matching 22

Remote Sensing Image Registration Via Feature Fusion and Enh...

引用

Proceedings of the 2022 6th International Conference on Electronic Information Technology and Computer Engineering

作者： Xianjing Li Ying Chen Wencheng Zhang Jiahao Wang Wei Wang School of Computer Science and Information Engineering Shanghai Institute of Technology China

ISBN: (纸本)9781450397148

image registration is a basic problem in image analysis and image processing. image registration has important applications in aerial image fusion, pattern recognition, three-dimensional reconstruction and other fields. Aiming at the problem of low registration accuracy and mismatching in remote sensing image registration, this paper proposes to use the Involution kernel to improve the ResNext network in the feature extraction stage and combines the SPANet attention mechanism with an improved ResNext network to improve the feature extraction ability of the network. In the feature matching stage, an enhanced matching method is proposed, which uses cross-correlation and nearest neighbor to second nearest neighbor ratio to filter out mismatched points to cope with complex images and background interference. The experimental results show that the proposed algorithm can achieve superior results in a variety of indexes compared with other algorithms, which proves that the proposed algorithm is effective.

关键词：

来源：评论

学校读者我要写书评

暂无评论

14th International Conference on Soft Computing and pattern recognition, SoCPaR 2022, and the 14th World Congress on Nature and Biologically Inspired Computing, NaBIC 2022

14th International Conference on Soft Computing and Pattern ...

引用

14th International Conference on Soft Computing and pattern recognition, SoCPaR 2022, and the 14th World Congress on Nature and Biologically Inspired Computing, NaBIC 2022

ISBN: (纸本)9783031275234

The proceedings contain 88 papers. The special focus in this conference is on Soft Computing and pattern recognition. The topics include: A Survey on Ensemble Multi-label Classifiers;semantic Type Detection in Tabular Data via Machine Learning Using Semi-synthetic Data;Handwritten Character recognition Using Deep LSTM Approach;two Phase Ensemble Learning Based Extractive Summarization for Short Documents;spider Monkey Based K-Means Dynamic Collaborative Filtering for Movie Recommendation Systems;Fuzzy Bi-GRU Based Hybrid Extractive and Abstractive Text Summarization for Long Multi-documents;a Genomic Signal processing-Based Coronavirus Classification Model Using Deep Learning with Web-Based Console;study on Drowsiness Detection System Using Deep Learning;convolutional Neural Networks for Crack Detection on Flexible Road Pavements;sexism Classification in Social Media Using Machine Learning Algorithms;improving the Performance of Classification via Clustering on the Students’ Academic Performance using Stacking Algorithm;deep Learning and Machine Learning-Based Lung Nodule Detection Systems – An Analysis;breast Cancer Classification Techniques – An Investigation;A Study on Student Performance Prediction and Intervention Mechanisms in MOOC;Bi-stage QWOA-Based Efficient Feature Selection for Enhanced Depression Detection Based on Facial Cues;some Cases of Prediction and Inference with Uncertainty;interpretable Approaches to Predict Evapotranspiration;deep Convolutional Neural Network Model for Classifying Alzheimer’s Disease;comparative Hybrid Deep Convolutional Learning Framework with Transfer Learning for Diagnosis of Lung Cancer;Review of Supply Chain Management Operational Resilience, Risk, and Disruption in COVID-19 Pandemic;adaptable Fog Computing Framework for Healthcare 4.0;ear images Classification Based on Data Augmentation and ResNeXt50;machine Learning Approaches for Crop Identification from remote sensing imagery: A Review;emotion recognition of Speech.

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：