Semantic Scholar Open Access 2024

Scenario recognition and tracking for cargo handling operations in autonomous and non-sparse outdoor industrial environments

J. V. D. Santos Guilherme Silva E. Borges P. Drews Silvia Silva da Costa Botelho

Abstrak

This paper presents a comprehensive study on the application of computer vision technologies for scenario recognition and tracking in industrial cargo handling operations, particularly within non-sparse and autonomous outdoor environments. We introduce a novel dataset and methodology aimed at real-time identification of various elements including personnel, containers, cages, equipment, boxes, and piping. Using the YOLOv8 neural network, our experiments achieved high accuracy, with precision reaching up to 87.3%. The model demonstrated effective detection and segmentation capabilities even in complex, non-sparse environments. These results suggest a significant enhancement in the decision-making processes and accident prevention strategies within industrial operations, underscoring the potential of advanced computer vision systems in improving safety and operational efficiency.

Topik & Kata Kunci

Penulis (5)

J

J. V. D. Santos

G

Guilherme Silva

E

E. Borges

P

P. Drews

S

Silvia Silva da Costa Botelho

Format Sitasi

Santos, J.V.D., Silva, G., Borges, E., Drews, P., Botelho, S.S.d.C. (2024). Scenario recognition and tracking for cargo handling operations in autonomous and non-sparse outdoor industrial environments. https://doi.org/10.1109/IECON55916.2024.10905130

Akses Cepat

Informasi Jurnal
Tahun Terbit
2024
Bahasa
en
Sumber Database
Semantic Scholar
DOI
10.1109/IECON55916.2024.10905130
Akses
Open Access ✓