People and Safety Belt Semantic Segmentation Dataset
The "People and Safety Belt Semantic Segmentation Dataset" is specifically curated for industrial applications, consisting of CCTV images captured within a factory environment at a resolution of 1920 x 1080 pixels. This dataset focuses on both instance and semantic segmentation, providing annotations for people and the seat belts they are wearing, aimed at enhancing safety compliance monitoring.
Sample



Specification
Dataset ID
MD-Image-054
Dataset Name
People and Safety Belt Semantic Segmentation Dataset
Data Type

Volume
About 1.5k
Data Collection
CCTY images from factory. Resolution is 1920*1080
Annotation
Semantic Segmentation,Instance Segmentation
Annotation Notes
The annotation of people and seat belts section.
Application Scenarios
Industry
Data Collections
This dataset includes around 1.5k images, each annotated to identify people and the seat belts they are using. The detailed segmentation assists in assessing safety protocol adherence in industrial settings, making it an essential tool for workplace safety monitoring, compliance verification, and automating the detection of safety equipment usage among workers.
Quality Assurance

Related products
Any further information, please contact us.