implemented face recognition with LBPH added

AttendanceApp/faceRecognition.py

+import cv2
+import os
+import numpy as np
+
+#This module contains all common functions that are called in tester.py file
+
+
+#Given an image below function returns rectangle for face detected alongwith gray scale imagess
+def faceDetection(test_img):
+    gray_img=cv2.cvtColor(test_img,cv2.COLOR_BGR2GRAY)#convert color image to grayscale
+    gray_img=cv2.normalize(gray_img, None, alpha=0, beta=1, norm_type=cv2.NORM_MINMAX, dtype=cv2.CV_32F)
+    gray_img=(255*gray_img).astype(np.uint8)
+    gray_img=cv2.fastNlMeansDenoising(gray_img);
+
+    face_haar_cascade=cv2.CascadeClassifier('HaarCascade/haarcascade_frontalface_default.xml')#Load haar classifier
+    faces=face_haar_cascade.detectMultiScale(gray_img,scaleFactor=1.32,minNeighbors=5)#detectMultiScale returns rectangles
+
+    return faces,gray_img
+
+#Given a directory below function returns part of gray_img which is face alongwith its label/ID
+def labels_for_training_data(directory):
+    faces=[]
+    faceID=[]
+
+    for path,subdirnames,filenames in os.walk(directory):
+        for filename in filenames:
+            if filename.startswith("."):
+                print("Skipping system file")#Skipping files that startwith .
+                continue
+
+            id=os.path.basename(path)#fetching subdirectory names
+            img_path=os.path.join(path,filename)#fetching image path
+            print("img_path:",img_path)
+            print("id:",id)
+            test_img=cv2.imread(img_path)#loading each image one by one
+            if test_img is None:
+                print("Image not loaded properly")
+                continue
+            faces_rect,gray_img=faceDetection(test_img)#Calling faceDetection function to return faces detected in particular image
+            if len(faces_rect)!=1:
+               continue #Since we are assuming only single person images are being fed to classifier
+            (x,y,w,h)=faces_rect[0]
+            roi_gray=gray_img[y:y+w,x:x+h]#cropping region of interest i.e. face area from grayscale image
+            faces.append(roi_gray)
+            faceID.append(int(id))
+    return faces,faceID
+
+def train_classifier(faces,faceID):
+    face_recognizer=cv2.face.LBPHFaceRecognizer_create()
+    face_recognizer.train(faces,np.array(faceID))
+    return face_recognizer
+
+#Below function draws bounding boxes around detected face in image
+def draw_rect(test_img,face):
+    (x,y,w,h)=face
+    cv2.rectangle(test_img,(x,y),(x+w,y+h),(255,0,0),thickness=4)
+
+#Below function writes name of person for detected label
+def put_text(test_img,text,x,y):
+    cv2.putText(test_img,text,(x,y),cv2.FONT_HERSHEY_DUPLEX,1,(255,0,0),3)
+
+#Save video frames
+def extractAndSaveFrames():
+    vidcap = cv2.VideoCapture('IT17098960.mp4')
+    success, image = vidcap.read()
+    count = 0
+    while success:
+        cv2.imwrite("frame%d.jpg" % count, image)  # save frame as JPEG file
+        success, image = vidcap.read()
+        print('Read a new frame: ', success)
+        count += 1
+
+
+
+
+
+
+
+
+
+

AttendanceApp/tester.py

+import cv2
+from .faceRecognition import faceDetection, train_classifier, labels_for_training_data, draw_rect, put_text
+
+
+#This module takes images  stored in diskand performs face recognition
+test_img=cv2.imread('TestImages/frame0.jpg')#test_img path
+faces_detected,gray_img=faceDetection(test_img)
+print("faces_detected:",faces_detected)
+
+
+#Comment belows lines when running this program second time.Since it saves training.yml file in directory
+faces,faceID=labels_for_training_data('trainingImages')
+face_recognizer=train_classifier(faces,faceID)
+face_recognizer.write('trainingData.yml')
+
+
+# For subsequent runs
+face_recognizer=cv2.face.LBPHFaceRecognizer_create()
+face_recognizer.read('trainingData.yml')#use this to load training data for subsequent runs
+
+name={0: "IT17138000", 1: "IT17100908", 2: "IT17098960"}#creating dictionary containing names for each label
+file = open("attendance.txt", "w")
+
+for face in faces_detected:
+    (x,y,w,h)=face
+    roi_gray=gray_img[y:y+h,x:x+h]
+    label,confidence=face_recognizer.predict(roi_gray)#predicting the label of given image
+    print("confidence:",confidence)
+    print("label:",label)
+    draw_rect(test_img,face)
+    predicted_name=name[label]
+    if(confidence>90):#If confidence more than 37 then don't print predicted face text on screen
+        continue
+    file.write(predicted_name)
+    put_text(test_img,predicted_name,x,y)
+file.close()
+resized_img=cv2.resize(test_img,(700,700))
+cv2.imshow("face dtecetion tutorial",resized_img)
+cv2.waitKey(0)#Waits indefinitely until a key is pressed
+cv2.destroyAllWindows
+
+
+
+
+