Generally:
Face Detection: Detect a face in the image. He seeks a common human face as a segment in the whole image. The output may be one or more. The output will be a rectangle or rectangles on the faces in the image. [Paul viola method]
Face Recognition: Recognize the input face from an already prepared database with the highest conformance score. The input must be one person, and the output will be the name, class name, or unknown person. [PCA, LDA]
source share