Human figure detection in Han portrait stone images via enhanced YOLO-v5

Heritage Science, Journal Year: 2024, Volume and Issue: 12(1)

Published: April 9, 2024

Abstract The unearthed Han Dynasty portrait stones are an important part of China’s ancient artistic heritage, and detecting human images in these is a critical prerequisite for studying their value. However, high-precision target detection techniques often result large number parameters, making them unsuitable portable devices. In this work, we propose new image model based on enhanced YOLO-v5. We discovered that the complex backgrounds, dense group targets, significant scale variations targets within scenes present challenges detection. Therefore, first incorporated SPD-Conv convolution Coordinate Attention self-attention mechanism modules into YOLO-v5 architecture, aiming to enhance model’s recognition precision small strengthen its resistance background disturbances. Moreover, introduce DIoU NMS Alpha-IoU Loss improve detector’s performance scenarios, reducing omission densely packed objects. Finally, experimental results from our collected dataset stone figure demonstrate method achieves fast convergence high accuracy. This approach can be better applied tasks special character backgrounds.

Language: Английский

---

Dense Buddha head object detection and counting YOLOv8 network based on multi-scale attention and data augmentation fusion

Published: Feb. 22, 2025

Language: Английский

---

Design and Implementation of ESP32-Based Edge Computing for Object Detection

Sensors, Journal Year: 2025, Volume and Issue: 25(6), P. 1656 - 1656

Published: March 7, 2025

This paper explores the application of ESP32 microcontroller in edge computing, focusing on design and implementation an server system to evaluate performance improvements achieved by integrating cloud computing. Responding growing need reduce burdens latency, this research develops server, detailing hardware architecture, software environment, communication protocols, framework. A complementary framework is also designed support processing. deep learning model for object recognition selected, trained, deployed server. Performance evaluation metrics, classification time, MQTT (Message Queuing Telemetry Transport) transmission data from various brokers are used assess performance, with particular attention impact image size adjustments. Experimental results demonstrate that significantly reduces bandwidth usage effectively alleviating load study discusses system’s strengths limitations, interprets experimental findings, suggests potential future applications. By AI IoT, demonstrates benefits localized processing enhancing efficiency reducing dependency.

Language: Английский

---

MODD-λ: Military Object Detection Dataset for Land-Air Integration and Cross-Domain Collaborative Unmanned Swarm Systems

Lecture notes in electrical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 590 - 598

Published: Jan. 1, 2025

Language: Английский

---

YOLO V8: An improved real-time detection of safety equipment in different lighting scenarios on construction sites

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 2, 2024

Abstract In this work we aimed to detect the safety equipment worn by workers on construction site using YOLOv8 model. Its a state-of-the-art deep learning model recognized for its speed and accuracy, in detecting objects within dynamic environments. Focusing classes such as Helmet, Vest, Gloves, Human, Boots, assess YOLOv8's efficacy real-time hazard detection. The have been labelled labelImg software training model, with that testing of different images videos carried out. After deploying trained it shows an impressive accuracy rate approximately 98.017% surpassing previous iterations. Additionally, our Recall Precision values achieve high levels at 94.9% 94.36% respectively, while F1 score mean Average (mAP) approximate 91% 91.9% respectively. These robust performance metrics underscore reliability effectiveness compared other existing YOLO models, marking significant advancement object detection management.

Language: Английский

---

Research on intelligent monitoring technology for roof damage of traditional Chinese residential buildings based on improved YOLOv8: taking ancient villages in southern Fujian as an example

Heritage Science, Journal Year: 2024, Volume and Issue: 12(1)

Published: July 4, 2024

Abstract In the process of preserving historical buildings in southern Fujian, China, it is crucial to provide timely and accurate statistical data classify damage traditional buildings. this study, a method based on improved YOLOv8 neural network proposed select aerial photographs six villages Xiamen Quanzhou cities Fujian Province as dataset, which contains total 3124 photographs. Based high-resolution orthophotographs obtained from UAV tilt photography, model was used make predictions. The main task first stage with value area, model's mAP (Mean Accuracy Rate) can reach 97.2% task. second uses segment images selected stage, detecting possible defects roofs, including collapses, missing tiles, unsuitable architectural additions, vegetation encroachment. segmentation task, reaches 89.4%, 1.5% improvement mAP50 (mean accuracy) compared original model, number parameters GFLOPs are reduced by 22% 15%, respectively. This effectively improve disease detection efficiency built heritage under complex terrain ground conditions.

Language: Английский

---

The YOLO Framework: A Comprehensive Review of Evolution, Applications, and Benchmarks in Object Detection

Computers, Journal Year: 2024, Volume and Issue: 13(12), P. 336 - 336

Published: Dec. 14, 2024

This paper provides a comprehensive review of the YOLO (You Only Look Once) framework up to its latest version, 11. As state-of-the-art model for object detection, has revolutionized field by achieving an optimal balance between speed and accuracy. The traces evolution variants, highlighting key architectural improvements, performance benchmarks, applications in domains such as healthcare, autonomous vehicles, robotics. It also evaluates framework’s strengths limitations practical scenarios, addressing challenges like small environmental variability, computational constraints. By synthesizing findings from recent research, this work identifies critical gaps literature outlines future directions enhance YOLO’s adaptability, robustness, integration into emerging technologies. researchers practitioners with valuable insights drive innovation detection related applications.

Language: Английский

---

Deployment and Verification of Custom Autonomous Low-Budget IoT Devices for Image Feature Extraction in Wheat

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 124636 - 124657

Published: Jan. 1, 2024

Language: Английский

---

YOLOv8-BCC: Lightweight Object Detection Model Boosts Urban Traffic Safety

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: April 5, 2024

With the rapid development of urbanization, role urban transportation systems has become increasingly prominent. However, traditional methods traffic management are struggling to cope with growing demands and complexity environments. In response this situation, we propose YOLOv8-BCC algorithm address existing shortcomings. Leveraging advanced technologies such as CFNet, CBAM attention modules, BIFPN structure, our aims enhance accuracy, real-time performance, adaptability intelligent detection systems. Experimental results demonstrate significant improvements in accuracy performance compared methods. The introduction provides a robust solution for enhancing safety management.

Language: Английский

---

Enhancing real-time instance segmentation for plant disease detection with improved YOLOv8-Seg algorithm

International Journal on Information Technologies and Security, Journal Year: 2024, Volume and Issue: 16(2), P. 27 - 38

Published: June 1, 2024

With widespread uses in areas as diverse traffic analysis and medical imaging, picture segmentation is a basic problem computer vision. Instance segmentation, which combines object recognition with powerful tool for item identification exact delineation. Using the Tomato Leaf disease dataset an example, this research delves into topic of training by capitalizing on simplicity enhanced YOLOv8-Seg models. leaf are focus instance-segmentation dataset, seeks to resolve pressing agricultural difficulties. One instance networks, YOLOv8n-Seg presented compared article purpose identification. The models tested difficult situations see how well they can detect separate garbage occurrences. Results show that useful agriculture accurately segmenting instances tomato detection.

Language: Английский

---