检索结果-内蒙古大学图书馆

image processing, Computer Vision and Machine Learning (ICICML), International Conference on

作者： Zhiyan Cui Na Lu Qian Wang Jingjing Guo Jiaming Yang School of Automation Science and Engineering Systems Engineering Institute Xi'an Shaanxi China Xi'an Jiaotong University Xi'an Shaanxi China Department of Construction and Real Estate School of Civil Engineering Southeast University Nanjing China College of Civil Engineering Hunan University Changsha China School of Civil Engineering Central South University Changsha China

ISBN: (纸本)9781665464697

Multiple object tracking is one of the critical directions in computer vision research. In the application of vision-based tracking methods, cameras are sometimes installed far from the targets to obtain a global view. There would be a large number of targets in the videos with relatively low resolution, which increases the difficulty of visual tracking. Applying existing tracking methods directly in such low-resolution scenarios will result in low recall and a large number of discontinued trajectory fragments, due to the instability of the target detection results. To alleviate the tracking performance degradation in low-resolution scenarios, a multiple object tracking method based on tracking compensation (MOT-TC) is proposed in this paper. A detector is applied to produce the candidate bounding boxes of the targets in the current frame. Then trajectories from previous frames are used to predict their states in the current frame. An assignment method is adopted to match the candidate bounding boxes to the predicted states. For the unmatched trajectories in the current frame, a single object tracking method for compensation is used to provide the target positions, which can increase the recall and reduce trajectory fragments. Meanwhile, a strategy based on the response map of single object tracking is designed to evaluate the tracking performance. Extensive experiments on low-resolution videos have shown that the proposed method outperforms the baseline and other state-of-the-art methods by a large margin.

关键词： Visualization Computer vision Target tracking Detectors Object detection Machine learning Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

Diagnosis of liquid-gas mixed sprays in the near-field region using femtosecond laser induced supercontinuum imaging method

引用

OPTICS EXPRESS 2020年第3期28卷 3298-3304页

作者： Wang, Mingxin Tan, Wenjiang Si, Jinhai Zheng, Yipeng Huang, Zhenqiang Xi An Jiao Tong Univ Key Lab Phys Elect & Devices Minist Educ Shannxi Key Lab Informat Photon TechSch Elect Sc Xianning Xilu 28 Xian 710049 Peoples R China

Direct femtosecond shadowgraphy and supercontinuum (SC)-illumination imaging methods for diagnosing liquid-gas mixed sprays in the near-field region of spray nozzles were compared. Some big spray structures can be captured using femtosecond shadowgraphy which can freeze the motion of the sprays. But the speckles caused by the interference of multi-scattered photons erode the edges of ligaments and conceal many fine droplets. SC-illumination imaging can not only freeze the motion of the sprays but also significantly suppressing the speckles, presenting a more realistic spray pattern. Based on the SC imaging technology, the effects of various factors such as flow ratio of gas to liquid (GLR), total flow and nozzle size on the spray were studied. (C) 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

关键词： Digital image processing image processing algorithms image quality Imaging systems Imaging techniques Speckle patterns

来源：评论

学校读者我要写书评

暂无评论

Analysis and Design of image Encryption algorithms Based on Interlaced Chaos 10th

Analysis and Design of Image Encryption Algorithms Based on ...

引用

10th International Conference on Computer Engineering and Networks, CENet 2020

作者： Li, Kangman Li, Qiuping College of Computer Science and Technology Hengyang Normal University Hengyang421002 China Hunan Provincial Key Laboratory of Intelligent Information Processing and Application Hengyang421002 China

ISBN: (纸本)9789811584619

For chaotic encryption algorithm, the algorithm of generating chaotic sequence is complex, and the data is float, which directly affects the speed of encryption. In this paper, an interlaced chaotic encryption algorithm is proposed, In this algorithm, the pixel interval is encrypted by chaotic sequence, and the chaotic sequences of other pixels are obtained by XOR the chaotic values of the front and back points, Experiments show that the algorithm can encrypt and decrypt images quickly and has good effect. © 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

Fast weighted CP decomposition for context-aware recommendation with explicit and implicit feedback

引用

EXPERT systems WITH APPLICATIONS 2022年 204卷

作者： Zhao, Jianli Zheng, Shidong Huo, Huan Gong, Maoguo Zhang, Tianheng Qu, Lijun Shandong Univ Sci & Technol Coll Comp Sci & Engn Qingdao Peoples R China Univ Technol Sydney Fac Engn & Informat Technol Ultimo Australia Xidian Univ Key Lab Intelligent Percept & Image Understanding Minist Educ Xi'an 710071 Peoples R China

As an effective method for high-dimensional information processing, tensor factorization has been widely used in context-aware recommendation. However, the multi-dimensional processing of tensor factorization usually leads to more serious data sparsity. Making full use of the data in recommendation systems, including a small amount of explicit rating data and a large amount of implicit feedback data such as clicks, plays, and purchases, can effectively alleviate this problem. Meanwhile, the processing of sparse explicit data and huge implicit feedback data makes tensor factorization face a severe challenge of computational efficiency. Therefore, this study proposes a fast tensor factorization model named FWCP that unifies explicit and implicit feedback information, which mainly studies as follows: (1) Assigning non-uniform weights to observed and unobserved values to unify explicit and implicit feedback information;(2) Memorizing intermediate data which generated by element-wise alternating least squares(ALS), and implementing parallel optimization of FWCP, which can reduce the training time of tensor factorization based on whole-data. The evaluation results on two real datasets showed that FWCP proposed in this study is superior to other context-aware recommendation algorithms.

关键词： Context-aware recommendation Tensor factorization Explicit and implicit feedback Element-wise alternating least squares Parallel optimization

来源：评论

学校读者我要写书评

暂无评论

基于PnP-ADMM SOC平台的单像素高分辨成像系统

引用

光学学报 2024年第16期44卷 133-143页

作者：黎淼张玲强王玺王晨燕陈朝锐郭兆辉赵雪吟重庆邮电大学光电工程学院重庆400065

针对目前单像素成像系统实时成像环境压缩采样重构的噪声鲁棒性需求,实现了基于即插即用的交替方向乘子法(PnP-ADMM)非凸反演框架的单像素成像硬件系统。针对所建立PnP-ADMM反演重构算法采用3种降噪算子完成数值仿真,仿真结果表明,所建... 详细信息

针对目前单像素成像系统实时成像环境压缩采样重构的噪声鲁棒性需求,实现了基于即插即用的交替方向乘子法(PnP-ADMM)非凸反演框架的单像素成像硬件系统。针对所建立PnP-ADMM反演重构算法采用3种降噪算子完成数值仿真,仿真结果表明,所建立单像素成像PnP-ADMM反演重构算法在采样率为12.50%及25.00%的条件下峰值信噪比可达24.82 dB及29.64 dB,结构相似性分别可达0.75及0.88。硬件系统实验结果表明,在采样率为25.00%的情况下,采用邻近算子、软阈值算子和全变分算子的PnP-ADMM片上系统(SOC)重构分辨率为256 pixel×256 pixel的图像的耗时分别为0.369 s、0.303 s和0.681 s,相比于仅使用ARM处理器分别提高了141.9倍、172.1倍和80.3倍。在采样率为25.00%、成像距离为300 cm条件下,所建立PnP-ADMM SOC单像素成像系统能够实现0.445~0.500 lp/mm的分辨率,具有较好的噪声抑制能力。

关键词：成像系统单像素成像交替方向乘子法即插即用 FPGA

来源：评论

学校读者我要写书评

暂无评论

FPGA Accelerated Real-time Recurrent All-Pairs Field Transforms for Optical Flow

FPGA Accelerated Real-time Recurrent All-Pairs Field Transfo...

引用

Chinese Automation Congress (CAC)

作者： Yingxiang Li Yingke Gao Zhiwen Su Shitao Chen Longjun Liu Institute of Artificial Intelligence and Robotics (IAIR) Xi’an Jiaotong University Xi’an China Beijing Institute of Control Engineering China Academy of Space Technology Beijing China

Optical flow algorithms based on deep learning have achieved excellent performance on multiple datasets, bringing new opportunity for optical flow estimation. Recurrent All-Pairs Field Transforms (RAFT) is one of the most powerful deep network based optical flow algorithms, but it is difficult to process in real time on the resource-limited embedded platform. In this paper, we propose RAFT-Lite by compressing the original RAFT model, which is more lightweight and suitable for hardware deployment. We further propose a hardware accelerating architecture on FPGA for RAFT-Lite, which provides an efficient scheduling strategy for the convolution in RAFT to achieve efficient pipeline and resource reuse. On xilinx ZCU102 evaluation board, the accelerated hardware system can reach 10.4fps processing images with a resolution of 512*396, which is 6.8x of i7-10700@2.90GHz and 46x of ARM Cortex-A53@1.50GHz. Besides, the power consumption is 13.103W.

关键词： image resolution Power demand image coding Convolution Transforms Hardware Real-time systems

来源：评论

学校读者我要写书评

暂无评论

Ensemble synthetic oversampling with pixel pair for class-imbalanced and small-sized hyperspectral data classification

引用

International Journal of Applied Earth Observation and Geoinformation 2024年 128卷

作者： Feng, Wei Long, Yijun Dauphin, Gabriel Quan, Yinghui Huang, Wenjiang xing, Mengdao Department of remote sensing science and technology School of Electronic Engineering Xidian University Xi'an 710071 China Xi'an Key Laboratory of Advanced Remote Sensing Xi'an 710071 China Key Laboratory of Collaborative Intelligence Systems Ministry of Education Xidian University Xi'an 710071 China Laboratory of Information Processing and Transmission L2TI Institut Galilée University Paris XIII France Key Laboratory of Digital Earth Science Aerospace Information Research Institute Chinese Academy of Sciences Beijing 100094 China Academy of Advanced Interdisciplinary Research Xidian University Xi'an 710071 China

Hyperspectral images (HSI) suffer from limited labeled data and the curse of dimensionality, which makes it difficult to classify imbalanced and small-sized HSI data. To address the mentioned issues, pixel pair features is designed to handle small-size problem. However, the augmented data inherits the class imbalance and even exhibits a larger imbalance ratio. Therefore, the ensemble-based Synthetic Minority Oversampling Technique (SMOTE) is integrated into pixel pair features to deal with imbalanced datasets and strengthen the quality of the augmented datasets. The proposed method is evaluated on five groups of HSI datasets and compared against seven state-of-the-art contrast algorithms, including ensemble-CNN (E-CNN), CNN with pixel pair features, ensemble CNN with pixel pair features, spatial–spectral E-ResNet, E-Transfering-VGGNet, ContextualNet, and CNN with K-means SMOTE. The proposed method demonstrates superior performance across three distinct difficulty levels, as determined by its average performance rank on five HSI datasets. This underscores its effectiveness in mitigating small size and class imbalance challenges. © 2024 The Authors

关键词： Ensemble learning Hyperspectral image classification Imbalanced data Small-sized data

来源：评论

学校读者我要写书评

暂无评论

Floating pollutant image target extraction algorithm based on immune extremum region

引用

DIGITAL SIGNAL processing 2022年第0期123卷 103442-103442页

作者： Yu, xiao Ye, xi Zhang, Shunxian Tianjin Univ Technol Tianjin Key Lab Control Theory & Applicat Complic Tianjin 300384 Peoples R China First Peoples Hosp Tancheng Tancheng Shandong Peoples R China

With the continuous development of human society, people's over-exploitation of nature leads to frequent environmental problems. A large number of floating objects appear on lakes, rivers, reservoirs and other water surfaces. Water floats have seriously damaged the ecological environment and directly threatened the survival and development of human beings. Therefore, for the sustainable development of human beings, we must solve the problem of water pollution. The detection of floating pollutants on water surface is the primary goal of water resource management. It is difficult to distinguish and identify water surface and pollutants when distortion and deformation occur due to illumination and other environmental factors. In this paper, based on the image of environmental water surface pollutants and the theory of region feature extraction based on artificial immune algorithm, a method of sewage floating object extraction based on immune extreme region is proposed to realize the recognition and detection of water surface pollutants. Algorithm of water pollutants through multi-scale Gaussian function image illumination correction, the congenital immune mechanism is adopted to the surface of the water pollutants at the beginning of the image segmentation, reduce the effects of other environmental factors, and then based on adaptive immunity immune extremum region affine invariability, recognition and detection figure floating pollutant distribution. Simulation results show that the proposed method is feasible. Compared with other traditional algorithms, the proposed method can effectively detect, identify and divide floating pollutants on water surface, which provides certain help for water resources management and protection. (C)& nbsp;2022 Elsevier Inc. All rights reserved.

关键词： Floating pollutant image Immune extremum region Pollutant image target extraction

来源：评论

学校读者我要写书评

暂无评论

基于关键点检测的红外弱小目标检测

引用

航空学报 2023年第10期44卷 284-294页

作者：王强吴乐天王勇王欢杨万扣东南大学自动化学院南京210096 江苏自动化研究所连云港222061 西北工业大学无人系统技术研究院西安710072 南京理工大学计算机科学与工程学院南京210094

红外弱小目标检测旨在从复杂的背景中检测出红外弱小目标,该技术在监视预警系统、精确制导等方面具有重要应用价值。针对已有的传统算法存在漏检、误检,深度学习中基于语义分割的检测方法易受“过分割”与“欠分割”影响等问题,提出了... 详细信息

红外弱小目标检测旨在从复杂的背景中检测出红外弱小目标,该技术在监视预警系统、精确制导等方面具有重要应用价值。针对已有的传统算法存在漏检、误检,深度学习中基于语义分割的检测方法易受“过分割”与“欠分割”影响等问题,提出了一种基于关键点检测的红外弱小目标检测算法(KeypointNet)。主要创新点为:直接优化目标中心点坐标,提高了检测效率,有效地保证了目标的检测率与虚警率;设计了一种从低层级到高层级的特征融合模块,获取了目标的多尺度信息,提升了检测效果。在相关数据集上的实验表明:KeypointNet算法的检测率能够达到97.58%,同时虚警率在2%以下,与其他算法相比取得了最好的效果。

关键词：红外弱小目标检测关键点检测深度学习 KeypointNet 图像处理

来源：评论

学校读者我要写书评

暂无评论

Data Dimension Reduction makes ML algorithms efficient

arXiv

引用

arxiv 2022年

作者： Khan, Wisal Turab, Muhammad Ahmad, Waqas Ahmad, Syed Hasnat Kumar, Kelash Luo, Bin School of Computer and Technology Anhui University Hefei230039 China Dept: Computer Systems Engineering Mehran Uni. of Enginerring & Technology Hyderabad Pakistan School of Aeronautics Northwestern Polytechnical University Shaanxi Xi’an710129 China Department Electrical Engineering Mehran Uni. of Enginerring & Technology Hyderabad Pakistan

Data dimension reduction (DDR) is all about mapping data from high dimensions to low dimensions, various techniques of DDR are being used for image dimension reduction like Random Projections, Principal Component Analysis (PCA), the Variance approach, LSA-Transform, the Combined and Direct approaches, and the New Random Approach. Autoencoders (AE) are used to learn end-to-end mapping. In this paper, we demonstrate that pre-processing not only speeds up the algorithms but also improves accuracy in both supervised and unsupervised learning. In pre-processing of DDR, first PCA based DDR is used for supervised learning, then we explore AE based DDR for unsupervised learning. In PCA based DDR, we first compare supervised learning algorithms accuracy and time before and after applying PCA. Similarly, in AE based DDR, we compare unsupervised learning algorithm accuracy and time before and after AE representation learning. Supervised learning algorithms including support-vector machines (SVM), Decision Tree with GINI index, Decision Tree with entropy and Stochastic Gradient Descent classifier (SGDC) and unsupervised learning algorithm including K-means clustering, are used for classification purpose. We used two datasets MNIST and FashionMNIST Our experiment shows that there is massive improvement in accuracy and time reduction after pre-processing in both supervised and unsupervised learning. Copyright © 2022, The Authors. All rights reserved.

关键词： Principal component 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：