检索结果-内蒙古大学图书馆

computer, vision and Intelligent Technology (ICCVIT), international conference on

作者： Yiren Qiu Jing Yang Ruping Shao Qilong Sha Nanjing Tech University Nanjing China

ISBN: (数字)9798331540043

ISBN: (纸本)9798331540050

Drone imagery analysis is an essential part of smart city construction and management. Object detection in UAV images has been a popular research topic recently. The images captured by drones usually have huge scale changes due to the flying and shooting at different altitudes, which puts a burden on traditional deep learning networks. The Transformer-based network has proven successful in computer vision tasks due to its strong long sequence memory analysis capabilities. This article will design a network for object detection and analysis of drone images based on the Transformer structure. The main challenge with using transformers is the high computational cost of processing image sequences, so the common method is to use local attention instead of global attention and a hierarchical structure to compensate for the loss of global vision. However, the large amount of information lost caused by using local attention can not be made up for with this action simply. This paper analyzes the ideas of multiscale and hierarchical structure concepts and attention mechanisms, and embeds multiscale concepts into the local attention structure to further make up for the global view information lost. In order to achieve this, we used the pooling concept to build multiscale pooling layers combined with a pyramid structure and applied it to the multi-head self-attention module. At the same time, we precut the image features entering the pooling block to reduce the length of the image token sequence. While reducing the calculation amount of the vision transformer, the global visual field information is increased as much as possible to obtain better contextual features. On VisDrone dataset, MRPVT as a backbone achieves 47.98% AP50 and 25.20% AP75 accuracy and on UAVDT’s car category achieves 98.84% AP50 and 89.27% AP75 accuracy. As expected, MRPVT delivers great performance on drone-captured images’ object detection and visual aesthetic assessment.

关键词： Deep learning computer vision Visualization Accuracy Smart cities Object detection Transformers Feature extraction image sequences Drones

来源：评论

学校读者我要写书评

暂无评论

Differential Oriented image Foresting Transform Segmentation by Seed Competition 2nd

Differential Oriented Image Foresting Transform Segmentation...

引用

2nd international conference on Discrete Geometry and Mathematical Morphology (DGMM)

作者： Condori, Marcos A. T. Miranda, Paulo A., V Univ Sao Paulo Inst Math & Stat BR-05508090 Sao Paulo SP Brazil

ISBN: (纸本)9783031198960;9783031198977

The image Foresting Transform (IFT) is a graph-based framework to develop image operators based on optimum connectivity between a root set and the remaining nodes, according to a given path-cost function. Oriented image Foresting Transform (OIFT) was proposed as an extension of some IFT-based segmentation methods to directed graphs, enabling them to support the processing of global object properties, such as connectedness, shape constraints, boundary polarity, and hierarchical constraints, allowing their customization to a given target object. OIFT lies in the intersection of the Generalized Graph Cut and the General Fuzzy Connectedness frameworks, inheriting their properties. Its returned segmentation is optimal, with respect to an appropriate graph cut measure, among all segmentations satisfying the given constraints. In this work, we propose the Differential Oriented image Foresting Transform (DOIFT), which allows multiple OIFT executions for different root sets, making the processing time proportional to the number of modified nodes. Experimental results show considerable efficiency gains over the sequential flow of OIFTs in image segmentation, while maintaining a good treatment of tie zones. We also demonstrate that the differential flow makes it feasible to incorporate area constraints in OIFT segmentation of multi-dimensional images.

关键词： Oriented image Foresting Transform image segmentation in directed graphs Generalized Graph Cut Differential algorithms

来源：评论

学校读者我要写书评

暂无评论

Abnormal Clustering and Cross Slicing Transformer for Insect Fine-Grained image Classification

Abnormal Clustering and Cross Slicing Transformer for Insect...

引用

Algorithm, image processing and Machine vision (AIPMV), international conference on

作者： Aokun Mei Hua Huo Big Data Analysis Laboratory of Henan Medical Big Data and Computing Intelligence Engineering Technology Research Center School of Information Engineering Henan University of Science and Technology Luoyang China

ISBN: (数字)9798350390254

ISBN: (纸本)9798350390261

Insect fine-grained image classification is an application scenario in fine-grained image classification. It not only has the characteristics of small inter-class differences and large intra-class differences, but also has the difficulty that some categories have multiple life-stage forms, which makes the general fine-grained image classification model difficult to play a role in insect scenes. To this end, based on the vision Transformer, we design a fine-grained classification network for insect images based on abnormal clustering and cross slicing, called ACCS-Trans. In the first stage of the model, we use segmentation and clustering operations to distinguish the special morphology of insects in those few-shot life stages. The model can avoid the interference of few-sample abnormal morphology on the class feature extraction of the current class during training. The second stage is the cross slicing module, which uses the anchor box of the image sample segmentation region in the first stage to cut the original sample image to form the main target image, which is used as the information supplement area of the original image. Finally, the image is divided into doubling patch groups by two vertical patch operations. In the third stage, we concatenate the multiplication patch group and input it into the vision Transformer network for class feature extraction. We fully experiment with ACCS-Trans on two insect image datasets. Compared with the current mainstream fine-grained image classification models, ACCS- Trans achieves state-of-the-art effects on both datasets, and we do ablation experiments for each module. The effect of each module on our ACCS- Trans is analyzed. The excellent performance of our ACCS- Trans in insect scenes is verified in these experiments, these provide new ideas for the task of fine-grained classification of insect images.

关键词： Training image segmentation computer vision Visualization Insects image edge detection Morphology Transformers Feature extraction image classification

来源：评论

学校读者我要写书评

暂无评论

image Talk: A Model for image Caption Generation with Voice

Image Talk: A Model for Image Caption Generation with Voice

引用

international conference on Futuristic Technologies (INCOFT)

作者： Akash Bhadange Rohan Bhole Vaishali Jabade Department of Electronics and Telecommunication Vishwakarma Institute of Technology Pune Pune India

The automatic creation of written descriptions from photographs represents a crucial point of interaction between computer vision and Natural Language processing (NLP) in the field of artificial intelligence (AI). This research paper presents an innovative approach to image caption generation employing a hybrid neural architecture that amalgamates Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. This novel methodology pushes the boundaries of image captioning techniques by incorporating cutting-edge techniques like Word2Vec embedding, ensuring the representation of semantic relationships between words in the generated captions. By leveraging the Xception model for robust feature extraction, this study advances the current state-of-the-art in image captioning methodologies. To quantify the efficacy of the proposed approach, comprehensive evaluations are conducted across diverse datasets including MS-COCO. These evaluations encompass rigorous assessment of both the accuracy of the generated captions in depicting the visual context and the fluency of the language employed. Performance metrics including BLEU (Bilingual Evaluation Understudy), METEOR, CIDER and ROUGE are utilized to quantitatively gauge the quality of the generated captions. The integration of a Text-to-Speech (TTS) feature for image captions is an additional enhancement to the proposed model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

FPGA-based Moving Object Detection Leveraging Frame Difference and Morphological Filtering

FPGA-based Moving Object Detection Leveraging Frame Differen...

引用

computer, vision and Intelligent Technology (ICCVIT), international conference on

作者： Yuexin Wang Xiaochen Jing Wenhao Wang Mengli Zhu Dongchen Zhu Bionic Vision System Laboratory Shanghai Institute of Microsystem and Information Technology Chinese Academy of Sciences Shanghai China University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9798331540043

ISBN: (纸本)9798331540050

Moving object detection plays a significant role in video surveillance. However, existing moving object detection methods often rely on software implementations, which means low real-time performance and high power consumption. This paper's core detection algorithm employs a frame difference method that is enhanced by morphological filtering. Additionally, we propose an architecture that integrates FPGA(Field Programmable Gate Array) and ARM(Advanced RISC Machine), fully leveraging the parallel computing advantages of FPGA and the high processing efficiency of ARM. The system utilizes a ZYNQ7000 SoC, coupled with an OV7725 camera for image capture and DDR3 SDRAM for data caching, to address the challenges of high-speed data processing and low power consumption. Experimental results show that the system meets the requirements for high real-time performance and low power consumption with a frame rate of 85.9375 frames per second and a total power consumption of 1.101 W.

关键词： Power demand Reduced instruction set computing Program processors Filtering Object detection Parallel processing Video surveillance Real-time systems Software Field programmable gate arrays

来源：评论

学校读者我要写书评

暂无评论

How to Transfer Strided Data into Accelerator

引用

AATCC JOURNAL OF RESEARCH 2021年第1_suppl期8卷 202-209页

作者： Gan, Xinbiao Wang, Zhiying Natl Univ Def Technol Sch Comp Changsha 410073 Peoples R China

Data transfer from a host central processing unit (CPU) into an accelerator is a performance bottleneck for applications accelerated by accelerators (such as general purpose digital signal processing (GPDSP), many integrated core (MIC), and general purpose graphics processing unit (GPGPU)). It is complicated and inefficient to transfer non-contiguous data with special respect to strided data. In this work, we present three approaches to transfer strided data for different scenarios: Redundant copy (RC), selective copy (SC), and transfer after transformed (TaT). We propose a space and time efficient method named TaT, in which strided data are transformed on the CPU first and then transferred into the accelerator. We simulated regions-of-interest (ROI) coding and validate proposed techniques. TaT was superior to RC on space efficiency and close to SC on saving space, but better than SC on time waste respectively.

关键词： Accelerator computer vision Redundant Copy Selective Copy Strided Data Transfer after Transform

来源：评论

学校读者我要写书评

暂无评论

Research and application of computer vision in the field of construction progress control

Research and application of computer vision in the field of ...

引用

Signal processing, Communications and Engineering Management conference (ISPCEM), international

作者： Zhilin Du Tianjin University of Technology Tianjin China

Project schedule lag is a problem often faced in engineering projects, and it is also an important reason for project cost overruns. It is of great significance to study the identification method of construction project image progress for monitoring and controlling project progress. The current project progress management mostly relies on manual inspection to collect on-site progress information, which is not only time-consuming and labor-intensive, but also the feedback of progress information is not timely. Therefore, this paper uses computer vision, deep learning and other technologies, combined with the logical reasoning of finite state machines, to study the identification method of construction project image progress at the construction process level.

关键词： Deep learning computer vision Analytical models Schedules image recognition Systematics Target recognition

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based image Analysis to Track Vehicles

Deep Learning-Based Image Analysis to Track Vehicles

引用

Disruptive Technologies (ICDT), international conference on

作者： P Prabakaran R. Jayasudha Mukta Sandhu Sandeep Gupta Department of Computer Applications School of Computing Sciences Vels Institute of Science Technology and Advanced Studies Chennai India Department of Mathematics Dr.N.G.P.Institute of Technology Coimbatore India Skill Department of Computer Science and Engineering Shri Vishwakarma Skill University Palwal Haryana India Department of Electrical Engineering Graphic Era (Deemed to be University) Dehradun India

There are many ways to track the traffic conditions on roads. With the rise of AI-based image processing technology, there has been a surge in interest in developing traffic monitoring systems that rely on camera vision data. An improved road monitoring system may be derived from the data collected in this research by placing a camera at a junction. The technique detects and classifies vehicles using a deep-learning-based methodology (YOLOv4). By comparing the positions of the vehicle with the map HD, lane-by-lane trajectories of the vehicle may be approximated. The predicted vehicle trajectories are used to calculate each lane's wait lengths and traffic volumes, broken down by travel direction. This technique was evaluated using five distinct metrics: vehicle detection rates, trajectory prediction, type of vehicle categorization, traffic volume estimate, and queue length assessment. Thousands of samples were used to assess its performance. Based on the findings, it is acceptable to assume that the system can identify vehicles with a 98% success rate and an error rate of less than 25% regarding lane-by-lane trip volume estimation and vehicle type classification. Therefore, this study's approach reveals the viability of collecting comprehensive traffic information from a camera set up at a junction. Current traffic monitoring schemes have a reasonable likelihood of being developed by this study's significant contribution, which is a strategy that combines AI with HD map approaches.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research on Robot image Recognition Based on Deep Learning

Research on Robot Image Recognition Based on Deep Learning

引用

Control, Electronics and computer Technology (ICCECT), IEEE international conference on

作者： He Zhang Lei Chen Xinyu Tong Yaxuan Li Huabo Guo Yizhou Yang Shixiang Lu College of Mechanical Engineering Tianjin University of Science & Technology Tianjin China College of Food Science and Engineering Tianjin University of Science & Technology Tianjin China College of Electronic Information and Automation Tianjin University of Science & Technology Tianjin China College of Artificial Intelligence Tianjin University of Science & Technology Tianjin China

ISBN: (数字)9798350380958

ISBN: (纸本)9798350380965

In recent years, the integration of image processing and machine vision technologies to achieve automation and intelligence has become a major focus of research. Intelligent management not only reduces labor and intensity but also enhances the efficiency of planting, harvesting, and picking crops and fruits. Initially, data is preprocessed by checking the dataset, removing erroneous labels and low-quality images, adjusting image size, transforming color space, denoising, and normalizing to enhance feature extraction and increase dataset size and diversity. This paper utilizes the YOLOV8 model to identify apples in images, obtaining bounding boxes and counting them. This count represents the number of apples. Simultaneously, a histogram is created to display the distribution of apple numbers based on these counts. Subsequently, the image dataset is read, images are displayed, and preprocessing and feature extraction functions are defined. The YOLOV8 model is loaded, position transformations are performed, and color and texture features are integrated. Finally, maturity is assessed, statistical analysis is conducted, and results are displayed. Lastly, the image dataset is read and an image is displayed. The paper defines preprocessing functions for binarization and morphological processing, contour extraction functions for apple detection and fitting, and functions for estimating apple area and quality. Statistical analysis and presentation of results are then performed.

关键词： Histograms Technological innovation Statistical analysis image color analysis Machine vision Noise reduction Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Real-Time Violence Detection and Alert System using MobileNetV2 and Cloud Firestore

Real-Time Violence Detection and Alert System using MobileNe...

引用

Networking and Communications (ICNWC), international conference on

作者： Mannu Thomas P Balamurugan Department of Networking and Communications SRM Institute of Science and Technology Kattankulathur Tamil Nadu

ISBN: (数字)9798350365269

ISBN: (纸本)9798350365276

The aggressive action greatly jeopardizes individual safety and general well-being. Various alternative tactics have been used to reduce violent behavior, including the installation and maintenance of surveillance systems. It is crucial for monitoring systems to automatically detect aggressive behavior and promptly provide warning signals to the authorities. computer vision and image processing have emerged as crucial study areas for identifying deviant and violent behavior, drawing in new researchers. The paper introduces a real- time violence detection and alert system that identifies violence in video streams by utilizing MobileNetV2. The system provides a scalable, precise, and adaptable public safety solution by using Cloud Firestore for data storage and administration, and a Telegram Bot for instant notifications.

关键词： Technological innovation Surveillance Scalability Streaming media Chatbots Real-time systems Public security

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：