PPE detection for Industrial Workers
- 0 Collaborators
By harnessing the capabilities of the oneDNN Intel oneAPI toolkit, this project aims to construct an AI-driven solution. It will analyse images or videos sourced from cameras or webcams, enabling robust detection of adherence to personal protective equipment protocols among construction site person. ...learn more
Project status: Under Development
oneAPI, Artificial Intelligence
Intel Technologies
oneAPI,
Other
Overview / Usage
This project focuses on revolutionising safety compliance in construction sites by leveraging AI and computer vision. It addresses the persistent challenge of ensuring proper personal protective equipment (PPE) usage among workers. By analyzing images or videos from cameras using AI algorithms, the system detects PPE presence accurately. This real-time monitoring minimizes safety risks and enhances compliance. The integration of the oneDNN Intel oneAPI toolkit optimizes AI performance. The project's outcomes offer automated, instant alerts for non-compliance and can be seamlessly integrated into site management systems. The research's adaptability extends its potential impact beyond construction, providing a safer work environment across industries.
Methodology / Approach
Our methodology involves a multifaceted approach that integrates advanced technologies to address the PPE compliance issue. We employ computer vision and AI algorithms to process visual data and determine the presence of safety gear. The project leverages the oneDNN Intel oneAPI toolkit to optimize AI performance, ensuring efficient processing of image or video inputs.
Technologies Used
The methodology employed in this project revolves around a fusion of AI, computer vision, and advanced technology to address the problem of personal protective equipment (PPE) compliance in construction sites.
1. Data Collection: Image and video data are captured through cameras or webcams placed strategically on construction sites.
2. Preprocessing: The collected data undergoes preprocessing to enhance quality, remove noise, and ensure consistency.
3. Computer Vision: Cutting-edge computer vision techniques, leveraging frameworks like OpenCV, are applied to analyze visual data. This involves object detection and classification to identify PPE items worn by workers.
4. AI Algorithms: Deep learning algorithms, implemented using tools like TensorFlow or PyTorch, are utilized for robust PPE detection. These algorithms are trained on labeled data to accurately identify safety gear.
5. Real-time Monitoring: The system is designed for real-time monitoring, continuously processing incoming data and providing immediate feedback on PPE compliance.
6. Integration of oneDNN and AI Analytics Toolkit: The oneDNN toolkit optimizes AI performance, enhancing accuracy and efficiency in PPE detection. AI analytics toolkit enhances data insights and decision-making, amplifying the project's impact and usability.
**7. Streamlit: **Streamlit is utilized for user-friendly deployment, providing interactive interfaces to visualize and interact with AI-based PPE detection results.