检索结果-内蒙古大学图书馆

19th International Conference on Ubiquitous Information Management and Communication, IMCOM 2025

作者： Isles, Melor Christian A. Pagkaliwangan, Dylan Lyle A. Caya, Meo Vincent C. School of Electrical Electronics and Computer Engineering Mapúa University Muralla St. Intramuros Manila1002 Philippines

ISBN: (纸本)9798331507817

The field of agriculture has slowly been integrating the use of technology with their daily operations to its advantage. Sunflower seeds, which are a popular mass-produced agricultural product, also employ these emerging techniques. This research aims to introduce a computer vision-based system designed to automate the classification and counting of sunflower seeds by leveraging YOLOv8 and deepsort algorithms. The system aims to identify and quantify three distinct seed varieties: Giant, Dwarf F1, and Mammoth Grey seeds using a Raspberry Pi-powered hardware setup. This setup was also used to capture high-quality images of the sunflower seeds which made up the dataset used. The researchers will utilize, along with the Raspberry Pi. a USB webcam to capture the video feed of the sunflower seeds and an LCD monitor to display the seed count and seed variety. Upon running multiple trials and utilizing a confusion matrix to analyse the system accuracy, the system was able to achieve an overall accuracy of 91.11% for its classification and 97.56% for its counting accuracy. © 2025 IEEE.

关键词： Computer Vision deepsort algorithm Raspberry Pi Sunflower Seeds YOLOv8

来源：评论

学校读者我要写书评

暂无评论

A sheep dynamic counting scheme based on the fusion between an improved-sparrow-search YOLOv5x-ECA model and few-shot deepsort algorithm

引用

COMPUTERS AND ELECTRONICS IN AGRICULTURE 2023年第1期206卷

作者： Cao, Yuanyang Chen, Jian Zhang, Zichao China Agr Univ Coll Engn Beijing 100083 Peoples R China Minist Agr & Rural Affairs Key Lab Smart Agr Technol Trop South China Guangzhou 510642 Peoples R China Minist Nat Resources Key Lab Urban Land Resources Monitoring & Simulat Shenzhen 518000 Peoples R China MNR Key Lab Spatial temporal Big Data Anal & Applicat Shanghai 200063 Peoples R China Beihang Univ State Key Lab Virtual Real Technol & Syst Beijing 100191 Peoples R China Jiangsu Univ Jiangsu Prov & Educ Minist Cosponsored Synergist I Ctr Modern Agr Equipment Zhenjiang 212013 Peoples R China

In order to improve the accuracy of sheep counting and avoid the interference of mutual occlusion caused by different moving speed among sheep, the concept of fusion between the improved YOLOv5x model based on the attention mechanism and deepsort algorithm is proposed. First, the ECA structure that is channel attention mechanism is used to optimize YOLOv5x model to strengthen the ability of capturing global information. Secondly, the sparrow search algorithm based on elite opposition-based learning strategy is used to optimize the learning rate of the detection model, so as to get the weight information of the optimal group to further improve the recognition rate of sheep. In the experiment, 800 high-resolution sheep images augmented by SRGAN network and data augmentation are used as model datasets, and the best weight information obtained by the YOLOv5x-ECA-SSA* model is used to accurately recognize sheep. According to the deepsort algorithm, the recognized sheep are tracked, predicted and matched optimally. The experimental results show that the test precision of YOLOv5x*, YOLOv5x-ECA* and YOLOv5x-ECA-SSA* based on the SRGAN and data enhancement to train are respectively 95.74%, 96.50% and 97.10%. The error rate of each model combined with deepsort algorithm to complete sheep dynamic counting is respectively 13%, 12% and 5%. Among them, the YOLOv5x-ECASSA* model has the highest mAP and best effect of sheep counting. The result can provide a new theorical method for realizing intelligent dynamic counting and tracking in the grazing process and provide a new technical application for intelligent animal husbandry.

关键词： YOLOv5x deepsort algorithm Sparrow Search algorithm Attention Mechanism Sheep Dynamic Counting

来源：评论

学校读者我要写书评

暂无评论

Engineering Vehicle Recognition and Multi-Object Tracking System based on YOLOv5 and deepsort algorithm 23

Engineering Vehicle Recognition and Multi-Object Tracking Sy...

引用

Proceedings of the 2023 5th International Conference on Video, Signal and Image Processing

作者： Yanbo Wu Xintao Liang Quan Cui Qingyan Wang Heilongjiang Government Affairs Big Data Center China School of Measurement-Control and Communication Engineering Harbin University of Science and Technology China School of Electronics and Information Engineering Harbin Institute of Technology China

ISBN: (纸本)9798400709272

This study explores traditional object detection algorithms and deep neural network-based engineering vehicle detection algorithms. We applied preprocessing algorithms such as image denoising, enhancement, and edge detection to process the images and constructed our own trainable dataset through labeling. Next, we used the YOLOv5 algorithm, which has high detection accuracy and real-time capability, for engineering vehicle detection. To address the issues of missed targets and predicted bounding box misalignment, we introduced the deepsort algorithm for target prediction and tracking. This algorithm utilizes Kalman filtering for estimation and updates and employs the Hungarian algorithm to associate data between consecutive frames, thereby achieving engineering vehicle tracking. To tackle the problem of frequent identity switching due to camera motion and non-uniform vehicle movements, we adopted the modified GIoU to calculate the intersection over union between trajectories and detected target bounding boxes, reducing identity jumps during the tracking process. Finally, we designed an engineering vehicle detection and tracking system and applied this algorithm to practical production.

关键词： deepsort algorithm YOLOv5 algorithm engineering vehicle detection multi-object tracking real-time detection

来源：评论

学校读者我要写书评

暂无评论

Multi-Camera Association Tracking algorithm for Pedestrian Target Based on Difference Image

引用

Systems and Soft Computing 2025年

作者： Shuai Ren College of Software Technology Henan Finance University Zhengzhou 450046 China

The current pedestrian target tracking algorithm (such as adjacent frame matching target tracking algorithm, deep learning YOLOv5 algorithm, etc) ignores pedestrian foreground image segmentation, resulting in significant errors in pedestrian target tracking and insufficient tracking results. Therefore, a multi-camera association tracking algorithm for pedestrians and targets based on differential images is designed. Multi-camera devices are used to collect pedestrian video sequence images, and the key frame difference image sample set is extracted. The initial background of the pedestrian image is modeled, and the foreground image is differentially segmented to construct the initial model of the differential image. The deepsort algorithm is used to complete the multi-pedestrian target association. The pedestrian target obeys the Laplacian random variable probability density function, and moves according to the center position of the bounding box to ensure that the target tends to move around the starting position, and realizes the multi-camera association tracking. The research method achieved maximum MOTA and MOTP values of 18.87% and 99.22% under different experimental times, demonstrating good association tracking ability. Moreover, the maximum comprehensive index of multiple pedestrian target association results approached 100%, while the minimum value far exceeded 95%. The tracking comprehensiveness and trajectory interruption rate of the research method were 98% and 1.2%, respectively, which were significantly better than other comparison algorithms. The processing speed reached 25FPS, effectively balancing computational efficiency. The experimental results verify that the proposed algorithm has ideal application effects.

关键词： Differential image Pedestrian targets Multi-camera device deepsort algorithm Association tracking algorithm

来源：评论

学校读者我要写书评

暂无评论

RGB-D Camera-Based Depth Measurement of Castings in Dynamic Environments

引用

INTERNATIONAL JOURNAL OF METALCASTING 2024年 1-14页

作者： Zhang, Long Chen, Zihao Miao, Jianhui Xie, Qian Hong, Jun Li, Hao Anhui Univ Technol Sch Met Engn Maanshan 243002 Anhui Province Peoples R China Anhui Highly Precis Casting Co Ltd Tech Dept Maanshan 238100 Anhui Province Peoples R China

In industrial production, the accuracy of depth measurement during casting gripping is often significantly compromised due to the influence of abnormal vibration, hindering the accurate positioning of the target casting. For this reason, this study proposes an RGB-D camera-based depth measurement method for castings in dynamic environments. The method initially achieves target point detection by embedding the gated differentiable image processing (GDIP) module into the YOLOv8 model and utilizes an enhanced deepsort algorithm to track the castings efficiently. Building upon this foundation, the statistically filtered point cloud depth data is fused with the inertial measurement unit (IMU) data processed by complementary filtering via an adaptive untraceable Kalman filter (AUKF) to ensure precise depth measurement results. The method is validated through simulations and field experiments, demonstrating superior robustness and accuracy compared to traditional depth measurement techniques in dynamic industrial environments. The simulation experiments evaluate the performance across varying vibration frequencies (10 Hz, 20 Hz, 40 Hz) and vibration amplitudes (1 mm, 2 mm, 5 mm). The results indicate that the system effectively mitigates interference while maintaining stable depth measurements under high-frequency and large-amplitude vibration conditions, significantly reducing relative error. In a real industrial environment, the depth measurement of a casting located 395 mm from the vibration source was improved from 4.757 to 2.336 mm, and the relative standard deviation (RSD) decreased from 0.24% to 0.12% through AUKF fusion. This study demonstrates that the depth measurement method proposed in this paper substantially enhances the robustness of depth measurement in heavy machinery vibration environments, ensuring stability and accuracy under real production conditions.

关键词： RGB-D camera depth measurement deepsort algorithm point cloud depth data IMU data adaptive untraceable Kalman filter

来源：评论

学校读者我要写书评

暂无评论

Automated productivity analysis of cable crane transportation using deep learning-based multi-object tracking

引用

AUTOMATION IN CONSTRUCTION 2024年 166卷

作者： Wang, Hao Yang, Qigui Liu, Quan Zhao, Chunju Zhou, Wei Zhang, Hongyang Liu, Jieyuan Wuhan Univ Inst Water Engn Sci Wuhan 430072 Peoples R China Changjiang Inst Survey Planning Design & Res Corp Wuhan 430010 Peoples R China Wuhan Univ Sch Water Resources & Hydropower Engn Wuhan 430072 Peoples R China Hubei Univ Technol Sch Civil Engn Architecture & Environm Wuhan 430068 Peoples R China

The automated monitoring of construction equipment productivity has been a crucial research topic in intelligent construction, supporting refined construction management. This paper presents a vision-based monitoring method for automated productivity analysis of cable crane transportation in dam construction. It employs a deep learning-based Multi-Object Tracking (MOT) method to track the moving trajectories of crane buckets. Based on the trajectory data, the transportation productivity of cable cranes is calculated accurately. The MOT method integrates small object detection layers, tracklet information (short trajectory fragments), and global position relationships into the YOLO-deepsort framework to enhance tracking performance in the construction industry. Experimental results show improvements of 95.9% in IDF1 and 92.1% in MOTA on three long videos collected from dam construction sites. These results indicate that the proposed method captures moving trajectories accurately and analyzes transportation productivity effectively.

关键词： Dam construction Multi-object tracking (MOT) Cable crane transportation deepsort algorithm YOLOv5 model Graph neural network (GNN)

来源：评论

学校读者我要写书评

暂无评论

Development of Real-Time Traffic Monitoring and Visualization System Using Stationary Roadside Sensor 22

Development of Real-Time Traffic Monitoring and Visualizatio...

引用

Proceedings of the 2022 11th International Conference on Networks, Communication and Computing

作者： Jundee Mark G. Molina Angelica S. Sotis Rogamestica C. Pascual James Earl D. Cubillas Junrie B. Matias College of Computing and Information Sciences Caraga State University Philippines

ISBN: (纸本)9781450398039

This study presented the development of a web-based system that visualizes real-time traffic by deploying lightweight and mobile monitoring devices at roadside intersections in the vicinity of Butuan City to assist commuters and drivers in making optimal decisions regarding efficient roadways for travel. The system can visualize the traffic situation at an intersection using IoT devices, wireless communication devices, and artificial intelligence. To analyze captured real-time traffic conditions and quantify traffic parameters using YOLOv5s and deepsort algorithm. deepsort algorithm is utilized for object tracking, whereas YOLOv5s is utilized for real-time object detection. The system has been successfully tested and is functional, providing a dashboard with real-time traffic information and a great platform for commuters and drivers in the vicinity of Butuan City to be aware of what is occurring on that particular roadway. This aids motorists, travelers, traffic and transportation offices, and organizations in obtaining traffic data for optimal decision-making, such as selecting the quickest route and boosting productivity and services. This would also aid in identifying specific transportation or traffic problems and generating cogent solutions.

关键词： YOLO Real-Time Traffic Monitoring deepsort algorithm

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：