检索结果-内蒙古大学图书馆

An underwater small target boundary segmentation method in forward- looking sonar images

APPLIED ACOUSTICS 2023年第1期207卷

作者： Zhang, Boming Zhou, Tian Shi, Zifeng Xu, Chao Yang, Kefan Yu, Xiaoyang Harbin Engn Univ Natl Key Lab Underwater Acoust Technol Harbin 150001 Peoples R China Harbin Engn Univ Key Lab Marine Informat Acquisit & Secur Minist Ind & Informat Technol Harbin Peoples R China Harbin Engn Univ Coll Underwater Acoust Engn Harbin 150001 Peoples R China Key Lab Polar Acoust & Applicat Minist Educ Harbin 150001 Peoples R China Yichang Testing Tech R&D Inst Yichang 443003 Peoples R China

Underwater target boundary segmentation is integral to forward-looking sonar image processing. However, small target boundary segmentation is always challenging due to the low resolution, low signal-to-noise ratio, and inhomogeneity of intensity of forward-looking sonar images. This paper pro-poses an improved level set segmentation method to accurately obtain the contour of small targets in the forward-looking sonar images: morphological reconstruction combined with the level set method (MRLSM). Compared with the classical level set and the level set method combined with the morpholog-ical method, MRLSM improves the accuracy and stability for forward-looking sonar image segmentation. Furthermore, the fuzzy C-means, Markov random field, and MRLSM are applied to the numerical simu-lations and experimental data for comparison. The deep learning methods are also used to compare the performances. The results demonstrated that the proposed method is more accurate, robust, and con-siderable for underwater small target boundary segmentation. (c) 2023 Elsevier Ltd. All rights reserved.

关键词： Level set method Morphological reconstruction Underwater small target segmentation Forward-looking sonar image

来源：评论

学校读者我要写书评

暂无评论

FEW SHOT LEARNING FOR INFRA-RED OBJECT RECOGNITION USING ANALYTICALLY DESIGNED LOW LEVEL FILTERS FOR data REPRESENTATION

FEW SHOT LEARNING FOR INFRA-RED OBJECT RECOGNITION USING ANA...

引用

IEEE International Conference on Image processing (ICIP)

作者： Arif, Maliha Mahalanobis, Abhijit Univ Cent Florida Ctr Res Comp Vis CRCV Orlando FL 32816 USA

ISBN: (纸本)9781665441155

It is well known that deep convolutional neural networks (CNNs) generalize well over large number of classes when ample training data is available. However, training with smaller datasets does not always achieve robust performance. In such cases, we show that using analytically derived filters in the lowest layer enables a network to achieve better performance than learning from scratch using a relatively small dataset. These class-agnostic filters represent the underlying manifold of the data space, and also generalize to new or unknown classes which may occur on the same manifold. This directly enables new classes to be learned with very few images by simply fine-tuning the final few layers of the network. We illustrate the advantages of our method using the publicly available set of infra-red images of vehicular ground targets. We compare a simple CNN trained using our method with transfer learning performed using the VGG-16 network, and show that when the number of training images is limited, the proposed approach not only achieves better results on the trained classes, but also outperforms a standard network for learning a new object class.

关键词： manifold eigen representation few shot learning sparse learning infra-red datasets

来源：评论

学校读者我要写书评

暂无评论

基于改进YOLOv8的凝视雷达小目标检测算法

引用

信号处理 2025年第5期41卷 853-866页

作者：周昶雯宋强张月中山大学电子与通信工程学院广东深圳518107 湖南科技大学计算机科学与工程学院湖南湘潭411201

小目标检测在低空飞行器管理、环境监测、边境安防等场景中发挥着至关重要的作用,是保障低空空间安全和推动低空经济发展的关键技术之一。现有的雷达目标检测算法在处理“低慢小”目标时,往往受到目标尺寸、信噪比(signal-to-Noise Rati... 详细信息

小目标检测在低空飞行器管理、环境监测、边境安防等场景中发挥着至关重要的作用,是保障低空空间安全和推动低空经济发展的关键技术之一。现有的雷达目标检测算法在处理“低慢小”目标时,往往受到目标尺寸、信噪比(signal-to-Noise Ratio,SNR)以及背景杂波等因素的限制,导致检测效果不佳。针对上述问题,本文提出将雷达距离-多普勒(Range-Doppler,RD)平面图输入改进后的YOLOv8模型,并结合实际雷达数据进行验证,实现了小目标检测性能的提升。本文利用全息凝视雷达采集了实测数据并进行数据标注和数据集制备。在模型设计方面,本文引入了高效通道注意力(Efficient Channel Attention,ECA)机制。ECA机制通过使用一维卷积代替传统的全连接层,在不增加额外参数量的前提下,有效地捕获了通道间的依赖关系。这种改进能够增强网络在不同通道上的特征选择能力,从而能够对小目标进行准确检测。同时,本文在模型中增加了小目标检测层,该检测层通过调整特征图的分辨率,使得模型能够对更小的目标进行特征提取和识别,从而弥补了常规检测网络对小目标处理能力不足的问题。此外,本文在实验部分选取了多种模型进行对比。实验结果表明,本文提出的改进模型在多个评价指标上均优于其他模型。与基准YOLOv8n相比,改进模型的精度提升了2.3%,mAP@0.5提升了1.9%,mAP@0.5-0.95提升了3.5%。总体来看,本文提出的模型在精度、召回率及mAP@0.5等指标上均表现出最佳的检测效果,验证了所提方法在雷达小目标检测中的有效性。

关键词：全息凝视雷达雷达小目标检测 YOLOv8 ECA注意力机制小目标检测层

来源：评论

学校读者我要写书评

暂无评论

Imaging Ultrasound Scattering targets using Density-Enhanced Chaotic Compressive Sampling

Imaging Ultrasound Scattering Targets using Density-Enhanced...

引用

Asia-Pacific signal and Information processing Association Annual Summit and Conference (APSIPA)

作者： Luong Thi Theu Tran Quang Huy Tran Thi Thuy Quynh Duc-Tan Tran Hoa Binh University Hanoi Vietnam Faculty of Physics Hanoi Pedagogical University 2 Hanoi Vietnam Faculty of Electronics & Telecommunications VNU University of Engineering & Technology Hanoi Vietnam Faculty of Electrical and Electronic Engineering Phenikaa University Hanoi Vietnam

The ultrasound imaging method is a highly promising technique used in various fields. The utilization of ultrasound technique's backscattering could also allow humans to observe objects as small as the size of the incident wavelength. However, the employing backscattering approach requires high computational calculations and poses a significant obstacle to the production of equipment using the ultrasonic method. In this work, we utilized the pseudo-random compression sampling method in the extended DBIM model, which takes tissue density factor. This is necessary because tissue density provides quantitative information about soft tissue and serves as an image contrast source. This method combines pseudo-random measurement and compressed sensing data recovery technique. The simulation results have proved the effectiveness of the proposed CCS-DBIM method. In terms of imaging quality, at compression ratios as low as 0.5, the large difference in imaging quality is evident. After 5 iterations, the normalization error of the DBIM and CCS-DBIM methods was 0.8828 and 0.5281 respectively, i.e. a 40% reduction in normalization errors when the number of measurements was only half of the variables. The disadvantage of the proposed CCS-DBIM method is that the imaging time is significantly greater than that of the DBIM method. However, with today's parallel computing technology, large amounts of data and computational time will no longer be a big problem.

关键词：

来源：评论

学校读者我要写书评

暂无评论

水下小目标合成孔径声呐层析成像技术研究

引用

声学技术 2023年第4期42卷 440-445页

作者：陈敬军曾赛海军装备部驻上海地区第七军事代表室上海201108 上海船舶电子设备研究所上海201108

水下小目标精细成像对于正确识别水下目标具有重要意义。目前,多波束成像声呐和条带合成孔径声呐是获取水下小目标图像的主要手段。水下目标的判别主要利用了目标图像的亮点特征,即使是同一目标从不同方位观测时得到的结果也可能差异较... 详细信息

水下小目标精细成像对于正确识别水下目标具有重要意义。目前,多波束成像声呐和条带合成孔径声呐是获取水下小目标图像的主要手段。水下目标的判别主要利用了目标图像的亮点特征,即使是同一目标从不同方位观测时得到的结果也可能差异较大,这给快速识别确认目标带来了困难。为解决该问题,提出了利用圆周合成孔径声呐对水下小目标进行水声层析成像信号处理方法,提高了声呐的多角度融合观测能力。仿真及试验数据处理结果表明,合成孔径声呐层析成像方法能够获得目标外形轮廓精细特征,有利于水下小目标的正确识别。

关键词：层析成像合成孔径目标识别

来源：评论

学校读者我要写书评

暂无评论

Photoacoustic Spatial Coherence Theory and Applications to Coherence-Based Image Contrast and Resolution

引用

IEEE TRANSACTIONS ON ULTRASONICS FERROELECTRICS AND FREQUENCY CONTROL 2020年第10期67卷 2069-2084页

作者： Graham, Michelle T. Bell, Muyinatu A. Lediju Johns Hopkins Univ Dept Elect & Comp Engn Baltimore MD 21218 USA Johns Hopkins Univ Dept Biomed Engn Baltimore MD 21218 USA Johns Hopkins Univ Dept Comp Sci Baltimore MD 21218 USA

The photoacoustic effect relies on optical transmission, which causes thermal expansion and generates acoustic signals. Coherence-based photoacoustic signal processing is often preferred over more traditional signal processing methods due to improved signal-to-noise ratios, imaging depth, and resolution in applications such as cell tracking, blood flow estimation, and imaging. However, these applications lack a theoretical spatial coherence model to support their implementation. In this article, the photoacoustic spatial coherence theory is derived to generate theoretical spatial coherence functions. These theoretical spatial coherence functions are compared with k-Wave simulated data and experimental data from point and circular targets (0.1-12 mm in diameter) with generally good agreement, particularly in the shorter spatial lag region. The derived theory was used to hypothesize and test previously unexplored principles for optimizing photoacoustic short-lag spatial coherence (SLSC) images, including the influence of the incident light profile on photoacoustic spatial coherence functions and associated SLSC image contrast and resolution. Results also confirm previous trends from experimental observations, including changes in SLSC image resolution and contrast as a function of the first M lags summed to create SLSC images. For example, small targets (e.g., <1-4 mm in diameter) can be imaged with larger M values to boost target contrast and resolution, and contrast can be further improved by reducing the illuminating beam to a size that is smaller than the target size. Overall, the presented theory provides a promising foundation to support a variety of coherence-based photoacoustic signal processing methods, and the associated theory-based simulation methods are more straightforward than the existing k-Wave simulation methods for SLSC images.

关键词： Acoustics Spatial coherence Imaging Apertures Spatial resolution Frequency control Acoustic signal processing coherence-based beamforming photoacoustic imaging spatial coherence

来源：评论

学校读者我要写书评

暂无评论

Learning Dynamic Local Context Representations for Infrared small Target Detection

arXiv

引用

arXiv 2024年

作者： Zhang, Guoyi Xu, Guangsheng Wang, Han Chen, Siyang Shan, Yunxiao Zhang, Xiaohu The School of Aeronautics and Astronautics Sun Yat-sen University Guangdong Shenzhen518107 China The School of Artificial Intelligence Sun Yat-sen University Guangdong Zhuhai519082 China The Southern Marine Science and Engineering Guangdong Laboratory Sun Yat-sen University Guangdong Zhuhai519082 China Shenzhen Institute Sun Yat-sen University Shenzhen510275 China The Guangdong Key Laboratory of Big Data Analysis and Processing SunYat-sen University Guangzhou510006 China

Infrared small target detection (ISTD) is challenging due to complex backgrounds, low signal-to-clutter ratios, and varying target sizes and shapes. Effective detection relies on capturing local contextual information at the appropriate scale. However, small-kernel CNNs have limited receptive fields, leading to false alarms, while transformer models, with global receptive fields, often treat small targets as noise, resulting in miss-detections. Hybrid models struggle to bridge the semantic gap between CNNs and transformers, causing high complexity. To address these challenges, we propose LCRNet, a novel method that learns dynamic local context representations for ISTD. The model consists of three components: (1) C2FBlock, inspired by PDE solvers, for efficient small target information capture;(2) DLC-Attention, a large-kernel attention mechanism that dynamically builds context and reduces feature redundancy;and (3) HLKConv, a hierarchical convolution operator based on large-kernel decomposition that preserves sparsity and mitigates the drawbacks of dilated convolutions. Despite its simplicity, with only 1.65M parameters, LCRNet achieves state-of-the-art (SOTA) performance. Experiments on multiple datasets, comparing LCRNet with 33 SOTA methods, demonstrate its superior performance and efficiency. Copyright © 2024, The Authors. All rights reserved.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Research on Spatial Angle Super-Resolution Algorithm Based on Adaptive Layered Sparse

Research on Spatial Angle Super-Resolution Algorithm Based o...

引用

Sensing, Measurement & data Analytics in the era of Artificial Intelligence (ICSMD), International Conference on

作者： Min Xue Yuexin Gao Zhixin Wu National Key Laboratory of Radar Signal Processing Xidian University Xi'an China School of Aerospace Science and Technology Xidian University Xi'an China

Compressed Sensing (CS) theory is based on the sparsity of signals and involves compressively sampling high-dimensional data to obtain a small number of linear observations that contain the complete information of the signals. By solving an optimization problem, the original signals can be recovered from these compressed linear observations. This paper combines CS theory with the sparse characteristics of targets in the spatial domain and proposes an adaptive layered sparse spatial angle super-resolution algorithm. This algorithm converts the target spatial angle super-resolution into the problem of using orthogonal basis to reconstruct sparse signals. Then, the target spatial angle super-resolution is carried out by using single snapshot data through optimization solution, and the target spatial angle domain is continuously reduced through layered orthogonal matching pursuit (LOMP) solution to complete the target spatial angle reconstruction. The effectiveness and robustness of the algorithm are verified by using the proposed algorithm to process the measured data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Biosensors Based Controller for small Unmanned Aerial Vehicle Navigation

Biosensors Based Controller for Small Unmanned Aerial Vehicl...

引用

SoutheastCon Conference

作者： McCoggle, Mylena Wilson, Shyra Rivera, Andrea Alba-Flores, Rocio Soloiu, Valentin Univ Michigan Dept Mech Engn Dearborn MI 48128 USA Kennesaw State Univ Dept Elect Engn Technol Kennesaw GA 30144 USA Interamer Univ Dept Mech Engn Bayamon PR USA Georgia Southern Univ Dept Elect Engn Statesboro GA USA Georgia Southern Univ Dept Mech Engn Statesboro GA USA

ISBN: (纸本)9781665406529

This work describes a system that uses electromyography (EMG) signals obtained from muscle sensors and an Artificial Neural Network (ANN) for signal classification and pattern recognition that is used to control a small unmanned aerial vehicle using specific arm movements. The main objective of this endeavor is the development an intelligent interface that allows the user to control the flight of a drone beyond direct manual control. The biosensors used in this work were the MyoWare Muscle sensors which contain two EMG electrodes and were used to collect signals from the posterior (extensor) and anterior (flexor) forearm, and the bicep. The collection of the raw signals from each sensor were performed using an Arduino Uno. data processing algorithms were developed with the purpose of classifying the signals generated by the arm's muscles when performing specific movements, namely: flexing, resting, arm-up, and arm motion from right to left. With these arm motions, roll control of the drone was achieved. MATLAB software was utilized to condition the signals and prepare them for the classification stage. To generate the input vector for the ANN and perform the classification, the root mean squared, and the standard deviation were processed for the signals from each electrode. The neuromuscular information was trained using an ANN with a single 10 neurons hidden layer to categorize the four targets. The result of the classification shows that an accuracy of 97.5% was obtained for a single user. Afterwards, classification results were used to generate the appropriate control signals from the computer to the drone through a Wi-Fi network connection. These procedures were successfully tested, where the drone responded successfully in real time to the commanded inputs.

关键词： Biosensors Electromyography Artificial Neural Network Arduino drone flight control Machine Learning

来源：评论

学校读者我要写书评

暂无评论

基于带标签CBMeMBer滤波器的低慢小群目标跟踪改进方法

引用

信号处理 2025年第5期41卷 906-923页

作者：钟展鸣宋强张月陈泽彬杨珺瑶中山大学电子与通信工程学院广东深圳518107 湖南科技大学计算机科学与工程学院湖南湘潭411201

群目标跟踪是实现无人机集群反制的关键步骤,具有重要的研究意义。相比传统的多目标跟踪算法,随机有限集(Random Finite Set,RFS)滤波在处理多目标数据关联与进行状态估计方面展现出显著的优势。然而在低慢小群目标的场景中,现有的RFS... 详细信息

群目标跟踪是实现无人机集群反制的关键步骤,具有重要的研究意义。相比传统的多目标跟踪算法,随机有限集(Random Finite Set,RFS)滤波在处理多目标数据关联与进行状态估计方面展现出显著的优势。然而在低慢小群目标的场景中,现有的RFS滤波方法普遍未考虑群特性对跟踪的影响,也未考虑群目标起始与航迹信息提取的问题。为此,本文结合了群建模、带标签势均衡多目标多伯努利(Cardinality Balanced Multi-target Multi-Bernoulli,CBMeMBer)滤波器以及自适应新生目标强度技术,提出一种针对低慢小群目标的跟踪方法。具体而言,本文首先通过虚拟领导-跟随者模型和无向图邻接矩阵对群目标结构进行建模;随后提出了一种三层优先级标签分配策略,改进了传统的带标签CBMeMBer滤波器,解决了标签冲突导致的轨迹混叠问题,提高了跟踪精度和算法运行效率;同时,设计了基于群目标场景与两点起始法的自适应新生目标强度算法,实现了RFS框架下群目标的自适应新生目标初始化;最后,仿真实验和基于全息凝视雷达的实测数据实验表明,所提方法在目标状态估计和轨迹质量方面表现优异,跟踪性能优于包括传统的带标签CBMeMBer滤波器在内的对比算法,且能有效避免轨迹交叉和混叠现象,充分展示了其在低慢小群目标精细化跟踪中的潜力和实际应用价值。

关键词：低慢小目标随机有限集群目标跟踪带标签势均衡多目标多伯努利滤波器全息凝视雷达

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：