The first step in analyzing an unknown image for faces is to perform face detection. The face detector searches through an image or video frame to locate regions that appear to contain a face. Media Server can then use the detected faces to perform face recognition or further face analysis.
Media Server returns the location of each face in the form of a bounding rectangle. You can configure face detection tasks to also determine the eye locations within a detected face. If this is enabled, Media Server returns the locations of the left and right eyes.
Possible applications of face detection include:
localizing and tracking a person
counting people
as a precursor to face recognition and analysis.
The performance of the face detector depends largely on the images that it processes. See Factors Affecting Face Detection Performance. You can also configure face detection for best performance.
|
|