python人脸识别应用环境搭建

工具DLIb+face_rcognition+opencv

安装过程如下：

step1：准备安装包

1.Anaconda3-5.0.1-Windows-x86_64.exe

2.dlib-19.7.0-cp36-cp36m-win_amd64.whl

3.face_recognition-1.0.0-py2.py3-none-any.whl

4.opencv_python-3.3.0.10-cp36-cp36m-win_amd64.whl

step2 安装

1.直接运行Anaconda安装文件

2.命令行下pip install dlib-19.7.0-cp36-cp36m-win_amd64.whl

3.命令行下pip install face_recognition-1.0.0-py2.py3-none-any.whl

4.命令行下pip install opencv_python-3.3.0.10-cp36-cp36m-win_amd64.whl

step3 实时人脸识别

spyder打开face_recognition\examples\facerec_from_webcam_faster.py，稍加修改即可实现。

代码如下：

import face_recognition
import cv2
import time
# This is a demo of running face recognition on live video from your webcam. It's a little more complicated than the
# other example, but it includes some basic performance tweaks to make things run a lot faster:
#   1. Process each video frame at 1/4 resolution (though still display it at full resolution)
#   2. Only detect faces in every other frame of video.


# PLEASE NOTE: This example requires OpenCV (the `cv2` library) to be installed only to read from your webcam.
# OpenCV is *not* required to use the face_recognition library. It's only required if you want to run this
# specific demo. If you have trouble installing it, try any of the other demos that don't require it instead.


# Get a reference to webcam #0 (the default one)
video_capture = cv2.VideoCapture(0)


# Load a sample picture and learn how to recognize it.
obama_image = face_recognition.load_image_file("hdk.jpg")
#128 dimensions feature
obama_face_encoding = face_recognition.face_encodings(obama_image)[0]


# Initialize some variables
face_locations = []
face_encodings = []
face_names = []
process_this_frame = True


while True:
    # Grab a single frame of video
    ret, frame = video_capture.read()
    
    #if(frame==None):  
        #continue


    # Resize frame of video to 1/4 size for faster face recognition processing
    small_frame = cv2.resize(frame, (0, 0), fx=0.25, fy=0.25)


    # Only process every other frame of video to save time
    if process_this_frame:
        # Find all the faces and face encodings in the current frame of video
        starttime = time.clock()
        face_locations = face_recognition.face_locations(small_frame)
        detectTime = time.clock()
        face_encodings = face_recognition.face_encodings(small_frame, face_locations)
        featureExTime = time.clock()
        print("face detect time:%f s" %(detectTime-starttime))
        print("face  features extract time:%f s" %(featureExTime-detectTime))
        face_names = []
        for face_encoding in face_encodings:
            # See if the face is a match for the known face(s)
            starttime =  time.clock()
            match = face_recognition.compare_faces([obama_face_encoding], face_encoding)
            fvTime= ( time.clock()-starttime)
            print("face verification time:%f s" %(fvTime))
            name = "Unknown"


            if match[0]:
                name = "hudekun"


            face_names.append(name)


    process_this_frame = not process_this_frame




    # Display the results
    for (top, right, bottom, left), name in zip(face_locations, face_names):
        # Scale back up face locations since the frame we detected in was scaled to 1/4 size
        top *= 4
        right *= 4
        bottom *= 4
        left *= 4


        # Draw a box around the face
        cv2.rectangle(frame, (left, top), (right, bottom), (0, 0, 255), 2)


        # Draw a label with a name below the face
        cv2.rectangle(frame, (left, bottom - 35), (right, bottom), (0, 0, 255), cv2.FILLED)
        font = cv2.FONT_HERSHEY_DUPLEX
        cv2.putText(frame, name, (left + 6, bottom - 6), font, 1.0, (255, 255, 255), 1)


    # Display the resulting image
    cv2.imshow('Video', frame)


    # Hit 'q' on the keyboard to quit!
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break


# Release handle to the webcam
video_capture.release()
cv2.destroyAllWindows()

step 4 效果如下：

总结，由于采用深度学习，人脸特征学习时间0.436052 s，需要做实时性改进。