Image Privacy Processing
Faces, license plates and sensitive text detected and masked with blur, mosaic or solid fill. The enterprise API validates samples first; you review the results.
What it does
Image and video redaction are available now; text and document redaction start with a sample assessment.
Faces
Several people in one frame are detected together, near and far.
Plates
Standard blue plates and new-energy green plates.
Sensitive text
Phone numbers, addresses and ID numbers printed in images; coverage validated on samples.
Three masking styles
Three outputs of the same frame, all real pipeline results (cropped).
Source
Same frame
Result
Blur: soft edges, natural look
Examples
Unredacted source on the left, real output on the right
Faces in a crowd
Near and distant faces all detected and blurred; clothing and background unchanged.
Street: faces and plates
Pedestrian faces and parked-car plates masked together; signs and shopfronts kept.
Traffic plates
Plates on several vehicles blurred one by one; lights and badges unchanged.
The crowd, street and traffic source images are AI-generated and show no real person; the results are real production output.
What affects results
Four things to know before integrating
Detection
Depends on sharpness, target size and occlusion; validate on your footage first.
Processing time
Asynchronous tasks; time depends on image size, content and queueing.
Styles
Blur, mosaic or solid fill.
Parameters
Masking style and strength are configured as agreed at integration.
API integration
Interface documentation and examples are provided after the trial application.
Process your images
For yourself, open Damaya; for your system, apply for an API trial.