Should your organisation be using manual or automated video redaction?
The need to redact video footage to protect privacy is one that many organisations are fully aware of. Whether you work in the public or private sector, thanks to GDPR, you will be required to blur faces or other sensitive information, such as licence plates and street names, in a video. There is software available that can help achieve this both manually and automatically, but which is the best for you?
What is the difference between manual and automated video redaction?
Manual video redaction involves somebody uploading a video to a video editing software and then personally reviewing it to select all that needs to be anonymised - faces, number plates, and other identifiable information. The user is required to review each piece of information and manually edit the blur, re-sizing or re-positioning it frame by frame.
This is a very time-consuming and can be a lot more costly. It is also subject to a lot more chance for human error. Even with shorter clips of video, the chance of missing faces that need to be blurred is much larger because of the sheer volume of video frames. This means a second opinion or secondary reviewer may be required to ensure privacy is upheld and the video redaction is successful.
The alternative is to automate this process.
Automated video redaction, such as Secure Redact, detects personal data in videos in order to redact it - with an accuracy rate of finding over 99% of personally identifiable information (PII) for redaction.
This means that for a 30-minute CCTV video that is a medium-busy scene, it would take 10 minutes to redact with Secure Redact. This is a lot quicker than the 12-17 days it would take if using manual video redaction techniques.
Automated redaction is the best option for firms handling a lot of sensitive data because it is simple to scale up - via SaaS and APIs - and accommodate large volumes of video.
Is manual or automated video redaction better?
Secure Redact is an automated solution that allows for the automated, anonymised video to be manually reviewed and edited, meaning you can tailor the redaction process to your needs. This could mean if you need to keep one individual visible but the rest of the video redacted, you can do so in our application.