Overview
The “AI Text” is an advanced feature that enhances user workflow by using AI to redact Personal Identifiable Information (PII) quickly. By using AI Text Automatic Redaction, you can detect and redact Street signs, license plates, and text in documents from your body-worn camera or CCTV footage in real-time!
For users who need to redact Street signs, license plates, and text in documents, CaseGuard’s AI Text Automatic Redaction can save hours’ worth of manual work. Even if you have to redact multiple different forms of text in a video, while leaving one street sign unredacted, we recommend running the AI first and then using the object detection panel to un-redact the street sign or text in question.
Follow the steps to automatically redact text within CaseGuard!
- In the left-hand Toolbar, select AI Video, the AI Automatic Detection window will appear with these options:
- Select the “Text” object then click “Next.”
- Apply the shape and Effect per object. The default is a Rectangle shape; add effects using the “Add Effect” dropdown. Selected Effect will appear in the Sample Preview image on the right.
- Fine-tune effect Level if necessary.
- Optionally, set Redaction Reason(s).
- Click “Next.”
- Select to apply your redactions to the entire video or a selected range. Applying only to a selected range will speed up the export process!
- Click “Next.”
- In the Audio Settings window, select if you would like to apply an audio redaction to your video.
- The selected audio effect will apply to the whole video.
- In the Advanced Options window, Change Confidence, Optimize Object Detection if the video is unsteady, optimize detection for 360 Videos, 4K, 8K, or Drone Footage, and/or Classify detection (if needed).
- Click “Next.”
- In the Summary window, Schedule for the system to run detection at a later time, and see a summary of your selected options.
- Begin the redaction process by clicking “Apply.”
- CaseGuard begins applying the selected effects to your video, you can see a progress bar in the bottom right.
- Detected objects are displayed as image thumbnails in the Object Detection Panel, as shown below:
- In the Object Detection Panel, you can manage redactions:
- To go to where the Object first appeared in the video, select the Go To Position .
- Click to view every frame the object was redacted in.
- Click to edit or delete the object.
- For individual selection, click the icon right of each object to Enable or Disable the effect on an object.
- Click the lock icon to lock an object from being changed.
- Click all the boxes beside the objects you would like to edit and click the Edit icon to update all the objects at once.
- Once you have reviewed your redaction, click “Export” from the Shortcuts Bar or the Task List Panel to process the file. The redacted file will appear in the Project Files Panel.
Let’s see it in action!