AI-Based Face Redaction in Law Enforcement Footage 

AI-Based Face Redaction in Law Enforcement Footage 

video redaction

This post is also available in: heעברית (Hebrew)

With the growing amount of image and video footage data documented by law enforcement organizations, officers have to process high video volume collected from body cameras and dashcams. Furthermore, in the US they have to respond to Freedom of Information Act requests by automating face redaction to preserve privacy, i.e. blurring out faces and personally identifiable markings in order to anonymize the footage to protect the privacy of people whose images are captured in stored video. Facial biometric data is typically very labor-intensive to anonymize manually.

An identity redaction tool that automates video and sensitive content editing has been developed by Panasonic Public Safety Solutions Division. The technology saves up to 90 percent of the time necessary to do the work manually, the company announced.

The new IDguard is based on  AI and machine learning technology that automate the video redaction process. Agencies can streamline file editing tasks and save time previously spent on uploading, storing, searching, editing, and sharing video content, according to

IDguard allows multiple video uploads for overnight redaction processing and multi-user management. It is compatible with MP4 videos, including from iPhones and Android devices, it has flexible configuration features for on-premises as well as cloud-based storage. 

The UK-based company Secure Redact also uses machine learning technology to provide fast and accurate redaction of security, survey and events footage. The technology provides automatic, selective anonymization of faces, number plates and personal objects. It can automatically blur all faces or select specific faces to blur/unblur. The technology is suitable for blurring images from body cams, dash cams, CCTV, passenger cam, drones, or smartphones, according to the company’s website.