检索结果-内蒙古大学图书馆

9th International Conference on Communication and Electronics Systems, ICCES 2024

作者： Manoj, N. Mohana Manu, A.S. Mallikarjun, N.Y. Ramakanth Kumar, P. Sreelakshmi, K. Rv College of Engineering® Bengaluru India Rv College of Engineering® Bengaluru India Rv College of Engineering® Bengaluru India

ISBN: (数字)9798350377972

ISBN: (纸本)9798350377972

In this paper YOLOv8 deep learning model is proposed for vehicle detection, classification, and counting for urban traffic surveillance applications on custom dataset. The model was trained with images and annotations done using the CVAT annotation tool, and it implemented in real-time using a mobile phone camera as a webcam via USB. The model accurately detected and classified four categories of vehicles such as two wheelers, three wheelers, four wheelers and large vehicles. A unique tracking ID approach was implemented to count the vehicles accurately. The obtained results prove the potential of custom model for vehicle detection and classification. The proposed model detects two wheelers, three wheelers, four wheelers and large vehicles with an accuracy of 93.6%, 86.4%, 88% and 89% respectively. © 2024 IEEE.

关键词： image annotation

来源：评论

学校读者我要写书评

暂无评论

Automatic Fish Species Identification Using Convolutional Neural Networks (CNN) 4

Automatic Fish Species Identification Using Convolutional Ne...

引用

4th International Conference on Sustainable Expert Systems, ICSES 2024

作者： Mary, I. Mettildha Ragunthar, T. Priyadharsini, M. Shyam Krishnaa, H. Sujit, M.K. Tamil Nadu Coimbatore India Srm Institute of Science and Technology Department of Computing Technologies Tamil Nadu Chennai India Vellore Institute Technology Department of Information Security Tamil Nadu Vellore India Siemens Healthineers Bangalore India Atheniac Bangalore India

ISBN: (纸本)9798331540364

Accurate identification of fish species is critical for applications in fisheries management., biodiversity monitoring., and conservation efforts. Conventional manual identification techniques take a lot of time and are prone to mistakes., particularly when working with big datasets and species that have similar physical characteristics. Automating tasks related to species identification has shown promise in recent breakthroughs in machine learning., especially in deep learning techniques like Convolutional Neural Networks (CNNs) and transfer learning. However., these systems face several challenges., including high variability in fish appearances., limited availability of labelled datasets., and the computational overhead required for real-time classification. This research study presents a fish species classification system using a transfer learning approach based on MobileNetV2. MobileNetV2., known for its efficiency and lightweight architecture., is fine-tuned on a custom fish dataset to enhance species identification accuracy while minimizing computational requirements. To increase the model's resilience and ability to generalise to new species., data augmentation approaches are used. The system is designed to deliver high classification accuracy while maintaining computational efficiency., making it suitable for deployment in real-world applications with limited resources. Experimental results demonstrate that the proposed system achieves significant improvements in accuracy compared to traditional methods., while also addressing the challenges of dataset variability and overfitting. © 2024 IEEE.

关键词： convolutional neural networks deep learning Fish species prediction image classification MobileNetV2 transfer learning

来源：评论

学校读者我要写书评

暂无评论

Spatial-channel transformer network based on mask-RCNN for efficient mushroom instance segmentation

引用

International Journal of Agricultural and Biological Engineering 2024年第4期17卷 227-235页

作者： Jiaoling Wang Weidong Song Wengang Zheng Qingchun Feng Mingfei Wang Chunjiang Zhao Northwest Agriculture and Forestry University Xi’an 712199China Nanjing Institute of Agricultural Mechanization Ministry of Agriculture and Rural AffairsNanjing 210014China Intelligent equipment Technology Research Center Beijing Academy of Agriculture and Forestry SciencesBeijing 100097China Zhejiang Provincial Key Laboratory of Agricultural Intelligent Equipment and Roboics/College of Biosystems Engineering and Food Science Zhejiang UniversityHangzhou 310058China

Edible mushrooms are rich in nutrients;however,harvesting mainly relies on manual *** localization of each mushroom is necessary to enable a robotic arm to accurately pick edible *** studies used detection algorithms that did not consider mushroom pixel-level *** these algorithms are combined with a depth map,the information is ***,in instance segmentation algorithms,convolutional neural network(CNN)-based methods are lightweight,and the extracted features are not *** guarantee real-time location detection and improve the accuracy of mushroom segmentation,this study proposed a new spatial-channel transformer network model based on Mask-CNN(SCT-Mask-RCNN).The fusion of Mask-RCNN with the self-attention mechanism extracts the global correlation outcomes of image features from the channel and spatial ***,Mask-RCNN was used to maintain a lightweight structure and extract local features using a spatial pooling pyramidal structure to achieve multiscale local feature fusion and improve detection *** results showed that the SCT-Mask-RCNN method achieved a segmentation accuracy of 0.750 on segm_Precision_mAP and detection accuracy of 0.638 on Bbox_Precision_*** to existing methods,the proposed method improved the accuracy of the evaluation metrics Bbox_Precision_mAP and segm_Precision_mAP by over 2%and 5%,respectively.

关键词： edible mushrooms picking instance segmentation deep learning algorithm

来源：评论

学校读者我要写书评

暂无评论

LIGHTWEIGHT DUAL-DOMAIN NETWORK FOR real-time MEDICAL image SEGMENTATION 29

LIGHTWEIGHT DUAL-DOMAIN NETWORK FOR REAL-TIME MEDICAL IMAGE ...

引用

IEEE International Conference on image processing (ICIP)

作者： Liu, Libo Fan, Xinxin Zhang, Xiaodong Hu, Qingmao Chinese Acad Sci Shenzhen Inst Adv Technol Shenzhen Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing Peoples R China Univ Chinese Acad Sci Beijing Peoples R China

ISBN: (数字)9781665496209

ISBN: (纸本)9781665496209

With the development of deep learning, deep convolution neural networks for medical image segmentation tasks have become more and more complex in pursuit of higher accuracy. In most scenarios, medical image segmentation pursues accuracy rather than speed, However, real-time performance is crucial in some scenarios, such as surgical navigation and diagnosis of acute stroke. So design of high-precision, lightweight and real-time medical image segmentation network has become an urgent need. To this end, a novel lightweight dual-domain network (LDD-Net) has been proposed in this paper. LDD-Net is comprised of two branches, learning respectively from the frequency domain and the spatial domain. In the frequency domain branch, the image spatial resolution is compressed via discrete cosine transform to have a large receptive field, so that better semantic context features can be learned. In the spatial domain branch, high-resolution feature representations with more details are learned. Finally, the learned features of these two branches are fused to yield high accuracy with low computational cost. The proposed method has been validated on two medical image segmentation datasets to yield the state-of-the-art performances with greatly reduced inference time and parameters of the learned models.

关键词： Medical image segmentation deep learning lightweight convolution neural networks lightweight dual-domain network

来源：评论

学校读者我要写书评

暂无评论

Casting-RCG and GRimage-Net: Generation of Casting Flaw Detection image and Defect Intelligent Detection 75

Casting-RCG and GRImage-Net: Generation of Casting Flaw Dete...

引用

75th World Foundry Congress, WFC 2024

作者： Hou, Mingjun Dong, Hao Ji, Xiaoyuan Li, Qinyang Wu, Chuhao Li, Shuohong Yin, Yajun Zhou, Jianxin State Key Laboratory of Materials Processing and Die & Mould Technology Huazhong University of Science and Technology Wuhan430074 China

The automatic defect detection of casting X-ray inspection images based on deep learning has the problems of single defect sample topography and poor network generalization. Therefore, the research on casting inspection image generation and multi-source image deep feature fusion defect detection is carried out. Firstly, a defect image simulation generation model based on RCG (Casting-RCG) was constructed, and a quasi-unconditional generation framework was built according to the defect features, which realized the high-quality generation of multiple defects such as shrinkage holes in Casting inspection images. Secondly, a defect localization and classification network based on deep feature fusion of generated images and real images (GRimage-Net) was created, and ablation experiments were carried out to verify the performance of the network. The results show that GRimage-Net based on Casting-RCG can significantly improve the accuracy, precision and recall rate of defect detection. It has better generalization than the single feature input model. The Casting-RCG and GRimage-Net created in this paper provide a new way to enrich the defect morphology and improve the generalization of the network, which has important theoretical guidance and engineering application value. © 2024 75th World Foundry Congress, WFC 2024. All rights reserved.

关键词： Precision casting

来源：评论

学校读者我要写书评

暂无评论

Adaptive Broad learning Network for LPI Radar Waveform Recognition 9

Adaptive Broad Learning Network for LPI Radar Waveform Recog...

引用

9th International Conference on Intelligent Computing and Signal processing, ICSP 2024

作者： Wang, Dong Cui, Tianshu Li, Ruike Huang, Yonghui Zhu, Yan National Space Science Center Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China China Academy of Aerospace Science and Innovation Beijing China Key Laboratory of Electronics and Information Technology for Space Systems Beijing China

ISBN: (纸本)9798350376548

Low probability of intercept (LPI) radar waveform recognition is a crucial branch in the field of electronic reconnaissance, serving as a vital method for acquiring information from non-cooperative radar sources. In recent times, deep learning models have demonstrated significant breakthroughs in the field of LPI radar waveform recognition (LWR). While these models significantly enhance recognition accuracy, the abundance of parameters and computations in the model renders training time-consuming, and the model structure lacks flexibility and scalability. To address these challenges, we propose an adaptive broad learning network (ABLN) for LWR. The ABLN transforms the time-frequency images (TFIs) of radar signals into feature nodes and enhancement nodes, with waveform types represented as output nodes. We further introduced a strategy for adaptively adjusting network parameters to reduce computational overheads, facilitating expedited optimization of the model structure. The simulation results show that ABLN has superior performance compared with other LWR methods. © 2024 IEEE.

关键词： image enhancement

来源：评论

学校读者我要写书评

暂无评论

Efficient Food image Segmentation Using YOLOv5: A Step Towards Automated Food Recognition 3

Efficient Food Image Segmentation Using YOLOv5: A Step Towar...

引用

3rd IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics, ICDCECE 2024

作者： Phadke, Rekha Chaurasia, Anshika Raj, Harsh Kumari, Nanshi Nitte Meenakshi Institute Of Technology Department Of Ece Bangalore India

ISBN: (纸本)9798350318609

Obesity rates have surged globally, necessitating effective weight management strategies. Documenting dietary intake is pivotal for weight loss management. This paper proposes a novel deep learning-based approach, Food image Segmentation, tailored for automated food image segmentation on custom datasets. Leveraging convolutional neural networks (CNNs) enhanced with transfer learning and attention mechanisms, our model achieves exceptional segmentation accuracy. Extensive experimentation validates its robustness and effectiveness, facilitating precise dietary assessment. Integrated into a user-friendly interface, our solution empowers individuals to capture food images and receive real-time nutritional analysis, aiding informed dietary choices. We address the challenge of deriving food information from images by proposing a CNN-based food image recognition algorithm. Our findings demonstrate significant advancements in automated food image segmentation, offering a promising avenue for personalized dietary management. We have achieved a mean average precision ranging from 0.407 to 0.895 in our work on various food instances. © 2024 IEEE.

关键词： image segmentation

来源：评论

学校读者我要写书评

暂无评论

An Efficient Information-Reinforced Lidar deep Completion Network without RGB Guided

引用

REMOTE SENSING 2022年第19期14卷 4689-4689页

作者： Wei, Ming Zhu, Ming Zhang, Yaoyuan Sun, Jiaqi Wang, Jiarong Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys Changchun 130033 Peoples R China Univ Chinese Acad Sci Beijing 100049 Peoples R China

Due to the sparsity of point clouds obtained by LIDAR, the depth information is usually not complete and dense. The depth completion task is to recover dense depth information from sparse depth information. However, most of the current deep completion networks use RGB images as guidance, which are more like a processing method of information fusion. They are not valid when there is only sparse depth data and no other color information. Therefore, this paper proposes an information-reinforced completion network for a single sparse depth input. We use a multi-resolution dense progressive fusion structure to maximize the multi-scale information and optimize the global situation by point folding. At the same time, we re-aggregate the confidence and impose another depth constraint on the pixel depth to make the depth estimation closer to the ground trues. Our experimental results on KITTI and NYU Depth v2 datasets show that the proposed network achieves better results than other unguided deep completion methods. And it is excellent in both accuracy and real-time performance.

关键词： depth completion lidar data processing image processing deep learning

来源：评论

学校读者我要写书评

暂无评论

Unsupervised deep learning enables real-time image registration of fast-scanning optical-resolution photoacoustic microscopy

引用

PHOTOACOUSTICS 2024年 38卷 100632页

作者： Hong, Xiaobin Tang, Furong Wang, Lidai Chen, Jiangbo South China Univ Technol Sch Mech & Automot Engn Guangzhou Guangdong Peoples R China City Univ Hong Kong Dept Biol & Chim Kowloon 83 Tat Chee Ave Hong Kong Peoples R China

A fast scanner of optical-resolution photoacoustic microscopy is inherently vulnerable to perturbation, leading to severe image distortion and significant misalignment among multiple 2D or 3D images. Restoration and registration of these images is critical for accurately quantifying dynamic information in long-term imaging. However, traditional registration algorithms face a great challenge in computational throughput. Here, we develop an unsupervised deep learning based registration network to achieve real-time image restoration and registration. This method can correct artifacts from B-scan distortion and remove misalignment among adjacent and repetitive images in real time. Compared with conventional intensity based registration algorithms, the throughput of the developed algorithm is improved by 50 times. After training, the new deep learning method performs better than conventional feature based image registration algorithms. The results show that the proposed method can accurately restore and register the images of fast-scanning photoacoustic microscopy in real time, offering a powerful tool to extract dynamic vascular structural and functional information.

关键词： Photoacoustic microscopy Unsupervised deep learning image registration

来源：评论

学校读者我要写书评

暂无评论

deep Q-Network for Cognitive Radar Anti-Jamming Strategy in Complex Electromagnetic Environment 12th

Deep Q-Network for Cognitive Radar Anti-Jamming Strategy in ...

引用

12th China Conference on Command and Control

作者： Liu, Chunhui Yuan, Yue Mao, Shuo Wang, Qiancheng Wang, Ben Nanjing Marine Radar Inst Nanjing Peoples R China Natl Key Lab Elect Effect & Secur Marine Equipmen Nanjing Peoples R China

ISBN: (纸本)9789819777723;9789819777709;9789819777693

The contemporary landscape of sophisticated and intelligent jamming techniques presents significant challenges for traditional radar anti-jamming methods, necessitating advancements in radar operational capabilities. To address the intricate and dynamic nature of jamming scenarios, alongside the limitations in performance assurance of manually crafted anti-jamming strategies and the suboptimal real-time responsiveness of radar systems, this study introduces an intelligent decision-making model founded on deep reinforcement learning (DRL). This model is meticulously structured, comprising a defined action space, state space, and reward function. Concurrently, the paper advocates a novel radar anti-jamming strategy learning approach based on the deep Q-Network (DQN), adept at mitigating external malicious interference. This approach enhances the integration efficiency and the doppler frequency resolution in radar echo processing. Comparative simulation outcomes affirm the superiority of the proposed intelligent decision model and training methodology over established methods like Proximal Policy Optimization (PPO) and Q-learning. Notably, the model demonstrates enhanced jamming suppression, robust generalization capabilities, accelerated response times, and a significant augmentation in the radar's autonomous decision-making prowess.

关键词： DQN Complex Electromagnetic Environment Cognitive Radar Anti-Jamming Decision

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：