Nowadays, face recognition is of utmost importance in almost every industry. But what if the situation demands its opposite? For privacy, identity protection and security, anonymising face by blurring it helps a lot.
This requirement can be accomplished by using OpenCV and Python.
- Plain blurring
- Pixelized blurring
- Perform face detection
- Extract face ROI
- Apply Gaussian Blur
- Replace the blurred face in original image / video frame
This is more demanding and pleasant to use.
- Detect and extract face ROI
- Divide it into MXN blocks
- Compute mean RGB pixel intensity of each block
- Annotate a rectangle on the block to create "pixelated" effect
- Apply the processed face ROI in original image
