Applying biometric identity is applied to face, in fact, two aspects: first, detecting a human face from an image or video frame, that is, a so-called "face detection"; second, for the detected The face is identified, and this face is, "Face Recognition). In terms of practical applications, it is a typical application in which a typical application is compared with fingerprints, iris identification, but not lost to a typical application in model identification, but at least a typical application in pattern recognition, at least a typical application of pattern recognition.
The following source code uses a hidden Markov model (HMM) to make face recognition, which is an application sample program for the OpenCV version 3.1, no longer included in version 4.0. So if you want to compile the source code, you need to install the OpenCV version 3.1. This version can be downloaded from SourceForg. About program use and algorithm description, refer to the page below (English):
Http://www.assuredigit.com/incoming/sourcecode/opencv/chinese_docs/Apppage/FaceRecognition/FaceRecognition.htm
And the paper:
Http://www.assuredigit.com/incoming/sourcecode/opencv/chinese_docs/papers/avbpa99.pdf
download link:
http://www.ssuredigit.com/program/hmmmdemo.ra