Artificial Intelligence for Routine Heritage Monitoring and Sustainable Planning of the Conservation of Historic Districts: A Case Study on Fujian Earthen Houses (Tulou)

Buildings, Journal Year: 2024, Volume and Issue: 14(7), P. 1915 - 1915

Published: June 22, 2024

With its advancements in relation to computer science, artificial intelligence has great potential for protecting and researching the world heritage Fujian earthen houses (Tulou) historical district. Wood is an important material used construction of (Tulou); wood both main structure buildings decoration. However, professionals must invest significant time energy evaluating any damage before repairing a building. In this context, study proposes optimizes detection method based on YOLOv8 model detecting wooden houses. Through multiple experiments adjustments, we gradually improved performance verified effectiveness reliability practical applications. The results are as follows: (1) This machine-learning-based object can efficiently accurately identify damaged contents, overcoming limitations traditional evaluation methods terms labor costs. approach will aid daily protection monitoring districts serves preliminary their renewal restoration. (2) rounds experiments, optimized significantly accuracy stability by removing samples with complex backgrounds, improving label quality, adjusting hyperparameters. final experiment, model’s overall mAP was only 57.00% at most. during field test, successfully identified nearly all points, including holes, stains, cracks analyzed building, effectively fulfilling requirements task. (3) KuiJu Lou test Tulou, also performed well environments able reliably detect types such structure. confirmed efficiency applications provided reliable technical support Tulou

Language: Английский

---

YOLOFLY: A Consumer-Centric Framework for Efficient Object Detection in UAV Imagery

Electronics, Journal Year: 2025, Volume and Issue: 14(3), P. 498 - 498

Published: Jan. 26, 2025

As an emerging edge device aimed at consumers, Unmanned Aerial Vehicles (UAVs) have attracted significant attention in the consumer electronics market, particularly for intelligent imaging applications. However, aerial image detection tasks face two major challenges: first, there are numerous small and overlapping objects that difficult to identify from perspective, second, if frame rate is not high enough, missed detections may occur when UAV moving quickly, which can negatively impact user experience by reducing accuracy, increasing likelihood of collision-avoidance failures, potentially causing unsafe flight behavior. To address these challenges, this paper proposes a novel YOLO (you only look once) framework, named YOLOFLY, includes C4f feature extraction module DWcDetect head make model lightweight, as well MPSA mechanism ACIoU loss function, improving accuracy performance consumer-grade UAVs. Extensive experiments on public VisDrone2019 dataset demonstrate YOLOFLY outperforms latest state-of-the-art model, YOLOv11n, 3.2% mAP50-95, reduces time 27.2 ms, decreases number parameters 0.6 M, cuts floating-point operations 1.8 B. Finally, testing real-world environments also yielded best results, including 3.75% reduction speeds. These findings validate superiority effectiveness YOLOFLY.

Language: Английский

---

Forestry Segmentation Using Depth Information: A Method for Cost Saving, Preservation, and Accuracy

Forests, Journal Year: 2025, Volume and Issue: 16(3), P. 431 - 431

Published: Feb. 27, 2025

Forests are critical ecosystems, supporting biodiversity, economic resources, and climate regulation. The traditional techniques applied in forestry segmentation based on RGB photos struggle challenging circumstances, such as fluctuating lighting, occlusions, densely overlapping structures, which results imprecise tree detection categorization. Despite their effectiveness, semantic models have trouble recognizing trees apart from background objects cluttered surroundings. In order to overcome these restrictions, this study advances management by integrating depth information into the YOLOv8 model using FinnForest dataset. Results show significant improvements accuracy, particularly for spruce trees, where mAP50 increased 0.778 0.848 mAP50-95 0.472 0.523. These findings demonstrate potential of depth-enhanced limitations RGB-based segmentation, complex forest environments with structures. Depth-enhanced enables precise mapping species, health, spatial arrangements, habitat analysis, wildfire risk assessment, sustainable resource management. By addressing challenges size, distance, lighting variations, approach supports accurate monitoring, improved conservation, automated decision-making forestry. This research highlights transformative integration models, laying a foundation broader applications environmental conservation. Future studies could expand dataset diversity, explore alternative technologies like LiDAR, benchmark against other architectures enhance performance adaptability further.

Language: Английский

---

Enriching the metadata of map images: a deep learning approach with geographic information systems-based data augmentation

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 181 - 203

Published: Jan. 1, 2025

Language: Английский

---

Fundamentals of Drone Navigation and the Role of Computer Vision

Advances in computational intelligence and robotics book series, Journal Year: 2025, Volume and Issue: unknown, P. 129 - 166

Published: March 28, 2025

Drone navigation is based on precise for efficient and secure performance delivery, surveillance, or rescue. Traditional GPS, inertial measurement units, magnetometers provides good guidance but inefficient in conditions with weakened signals unpredictable obstacles. Computer vision changing this. By equipping drones to perceive understand visual information about their surrounding space, it makes decision-making independent, allows better past obstacles, builds real-time maps. Object detection, optical flow, SLAM are some techniques being applied aerial robotics today. Vision complemented enhanced when combined other sensors like LiDAR making feasible complex terrains. However, processing high volumes of data remains a challenge. Advances edge computing AI-driven perception helping overcome these limitations, bringing faster more onboard processing.

Language: Английский

---

Advanced Object Detection for Maritime Fire Safety

Fire, Journal Year: 2024, Volume and Issue: 7(12), P. 430 - 430

Published: Nov. 25, 2024

In this study, we propose an advanced object detection model for fire and smoke in maritime environments, leveraging the DETR (Detection with Transformers) framework. To address specific challenges of shipboard detection, such as varying lighting conditions, occlusions, complex structure ships, enhance baseline by integrating EfficientNet-B0 backbone. This modification aims to improve accuracy while maintaining computational efficiency. We utilize a custom dataset images captured from diverse incorporating range data augmentation techniques increase robustness. The proposed is evaluated against YOLOv5 variants, showing significant improvements Average Precision (AP), especially detecting small medium-sized objects. Our achieves superior AP score 38.7 outperforms alternative models across multiple IoU thresholds (AP50, AP75), particularly scenarios requiring high precision occluded experimental results highlight model’s efficacy early demonstrating its potential deployment real-time safety monitoring systems. These findings provide foundation future research aimed at enhancing challenging environments.

Language: Английский

---

Enhanced Detection and Classification of Microplastics in Marine Environments Using Deep Learning

Regional Studies in Marine Science, Journal Year: 2024, Volume and Issue: unknown, P. 103880 - 103880

Published: Oct. 1, 2024

Language: Английский

---

The use of convolutional neural networks for abnormal behavior recognition in crowd scenes

Information Processing & Management, Journal Year: 2024, Volume and Issue: 62(1), P. 103880 - 103880

Published: Sept. 10, 2024

Language: Английский

---

Metaverse: Fission the architecture

International Journal of Science and Research Archive, Journal Year: 2024, Volume and Issue: 12(1), P. 446 - 470

Published: May 12, 2024

Modern technology comes with new opportunities and difficulties since it is constantly evolving. The that has generated the utmost buzz intrigue in recent years Metaverse. Although Metaverse not an entirely word, attracted more attention because Facebook changed its name to Meta. However, despite enormous interest prospects, still needs be determined how ethical issues will addressed users' privacy protected system. Furthermore, system must earn confidence acceptance of user by fulfilling main criteria Trustworthy AI. Therefore, this paper focused on making trustworthy. This covered Metaverse's history, essential elements, current business market, future opportunities, challenges. Further, manuscript discussed pillars trustworthy AI, factors, way trustworthiness. Finally, combined these concepts identified elements contributing credibility.

Language: Английский

---

Multimodal geometric AutoEncoder (MGAE) for rail fasteners tightness evaluation with point clouds & monocular depth fusion

Measurement, Journal Year: 2024, Volume and Issue: unknown, P. 116557 - 116557

Published: Dec. 1, 2024

Language: Английский

---