SAFR measures the sharpness and contrast with each frame of input video. This information is used to evaluate the quality of the input image to determine if it’s acceptable for recognition or learning. SAFR Recognition preferences allows you to configure the thresholds SAFR uses to determine if an image should be used for recognition or if the images is of sufficiently high quality to enroll.
Sharpness and contrast are reported as a decimal value ranging from 0 (very low quality) to 1.0 (high quality).
SAFR reports the value of sharpness and contrast in the Recognition Details view. Below is an example of SAFR Recognition Details panel with Center Pose Quality of 0.46, sharpness (S) reported as 0.76 and contrast (C) reported as 0.69.
This information can then be used to adjust Recognition settings in SAFR Preferences dialog to include or exclude subjects from recognition:
