Overview
Ensuring safety is a top priority globally, especially for those working in environments like construction sites, labs, and medical practitioners. Despite implementing sufficient controls, workers still face safety hazards. In construction and healthcare, the use of Personal Protective Equipment (PPE) is crucial.
VIDIZMO provides a solution driven by advanced technology that can detect workers who have forgotten to wear their protective equipment. This ensures that every individual on the site is protected from potential hazards. VIDIZMO's feature of detecting PPE in a video aims to improve workplace safety by accurately detecting the presence of PPE in videos.
Concept
To identify where individuals in videos are not wearing Personal Protective Equipment (PPE), users can leverage VIDIZMO to extract such information regarding the following Personal Protective Equipment (PPE) such as Mask, Protective Hats, and Hand Covers (such as surgical gloves).
VIDIZMO utilizes Amazon Rekognition to identify Personal Protective Equipment (PPE) worn by individuals in videos. Enabling the AWS Indexer application is required to use this feature. VIDIZMO has seamlessly integrated Amazon Rekognition for PPE detection.
In VIDIZMO, the PPE feature works for videos. The system achieves this by extracting frames at a customizable frequency, allowing users to specify up to a range of 1-10 frames. The system processes these frames per second to detect instances of PPE non-compliance. In a video to identify individuals not wearing PPE, the detection response includes the following details:
Person detection in the video.
Body parts where PPE is not worn, such as face, head, left hand, and right hand.
Types of PPE are not detected on specific body parts, such as face cover, hand cover, and head cover.
Bounding boxes indicating the locations of persons and PPE items are not detected in the video.
When configuring the AWS indexer, the user can customize the Minimum Track Detection Duration, which starts with a minimum value of 1 second and can be further customized based on specific requirements. This duration, measured in seconds, is utilized by the system to track individuals in a video. Individuals with a tracked duration exceeding this specified threshold are subjected to Personal Protective Equipment (PPE) checks in the video. The system uses this defined duration to monitor and assess individuals for PPE compliance.
Types of PPE Filter
The user can detect Personal Protective Equipment (PPE) in videos and filter content based on specific labels. The system distinguishes various aspects:
- Bare Face Detection: This feature identifies common face coverings such as surgical masks, N95 masks, and cloth masks. The system alerts when an individual without a face cover is detected in a video.
- No Left Glove and No Right Glove Detection: The system is equipped to recognize hand covers like surgical gloves and safety gloves. It filters individuals who are not wearing these gloves in the video.
- Bare Head Detection: This functionality detects the presence of hard hats and helmets, filtering out individuals in the video who are not wearing them.
In the above image, it's important to observe the absence of head cover and hand cover. The system promptly detects and notifies that the individual is not wearing Personal Protective Equipment (PPE) on the head and hands. In this specific scenario, all filter labels are activated. If any filter is not selected, the system won't identify non-compliance with PPE in the video for that particular aspect.
Up Next
This functionality empowers VIDIZMO users to enhance workplace safety protocols. This capability serves as a valuable tool for ensuring compliance with safety regulations and promoting a secure working environment.
To learn more about how to use this feature, refer to our next article, "How to Detect PPE in VIDIZMO."