webcam.py

import os
import cv2
import sys
import numpy as np
import face_recognition 
import csv
import datetime

def main():
    #Function definition to encode images
    def encode(images):
        #List of Encoded Images and a list to capture images with faces missing
        encoded_list = []
        missing = []
        #Function to encode each image in the image list
        for image, name in zip(images, employee_names):
            image = cv2.cvtColor(image,cv2.COLOR_BGR2RGB)
            try:
                encode = face_recognition.face_encodings(image)[0]
                encoded_list.append(encode)
            except:
                print("[-] No face found in image {0}!".format(name))
                missing.append(name)
        #Return encoded image list
        return encoded_list, missing
    #Function to make the attendance log file
    def createcsv():
        with open('logs.csv', 'w') as csvfile:
            filewriter = csv.writer(csvfile, delimiter=',',
                                    quotechar='|', quoting=csv.QUOTE_MINIMAL)
            filewriter.writerow(['Name', 'Time of Entry', 'Date'])
    #Function to mark each employee's entry time and not repeat it once the are marked to prevent errors
    def attendance(name):
        with open('logs.csv', 'r+') as csvfile:
            data = csvfile.readlines()
            name_list = []
            for line in data:
                entry = line.split(",")
                name_list.append(entry[0])
            if name not in name_list:
                now = datetime.datetime.now()
                time = now.strftime("%H:%M:%S")
                date = now.strftime(r"%d:%m:%Y")
                csvfile.writelines("\n{0},{1},{2}".format(name,time,date))

    path = os.path.dirname(__file__)+"/images/"
    #Exception if images folder doesn't exist
    if(os.path.isdir(path)):
        pass
    else:
        print("Please create the 'images' folder inside the script directory.")
        sys.exit(1)
    images = []
    employee_names = []
    lst = os.listdir(path)
    #Return exception if no images are present
    if not lst:
        print("Please add some employee images for the model to recognize.")
        sys.exit(1)
    #Looping through each employee name in list
    for name in lst:
        #Removes the extension from the names and updates Employee Name List
        employee_names.append(os.path.splitext(name)[0])
        #Images loaded 
        current_image = face_recognition.load_image_file("{0}{1}".format(path, name))
        #Adds each image to the image list for the AI to work with
        images.append(current_image)

    print("[+] Please wait initializing.......")

    #List of encoded image of employees created
    encoded_employee_list, missing = encode(images)
    #Sanitize the final list
    for name in missing:
        employee_names.remove(name)

    print("[+] Finished Initialization!")

    #Create the log file
    createcsv()

    #Video capture through webcam initiated
    capture = cv2.VideoCapture(0)
    #Loop to capture images of employees frame by frame via a webcam which is of the device id 0 here
    print("\n[+] If you want to terminate the live feed and exit, just press ctrl+c")
    while True:
        ret, image = capture.read()
        try:
            #Resize image for faster processing
            small_image = cv2.resize(image, (0,0), None, 0.20, 0.20)
            #Find face locations in webcam feed
            current_face = face_recognition.face_locations(small_image)
            #Encode found faces
            current_encoding = face_recognition.face_encodings(small_image, current_face)
            #Loop to iterate through all the found faces and their encoding in current frame
            for encoded_face, face_location in zip(current_encoding, current_face):
                #Compare and match encoded faces from feed frame by frame to known encoded list of employee images
                match = face_recognition.compare_faces(encoded_employee_list, encoded_face)
                #Compare face distance values from live feed with values from encoded employee list
                face_distance = face_recognition.face_distance(encoded_employee_list, encoded_face)
                #Select the employee index for the employee with the highest similarity with face in current frame
                match_index = np.argmin(face_distance)
                #Condition to draw a rectangle and display the name of the found face in feed 
                if match[match_index]:
                    name = employee_names[match_index].title()
                    top, right, bottom, left = face_location 
                    top*=5
                    bottom*=5
                    right*=5
                    left*=5
                    #Draw Rectangle on face 
                    cv2.rectangle(image, (left, top), (right, bottom), (0,0,255), 1)
                    cv2.rectangle(image, (left, bottom-15), (right, bottom), (0,0,255), cv2.FILLED)
                    #Put name on text
                    cv2.putText(image, name, (left+3, bottom-3), cv2.FONT_HERSHEY_PLAIN, 1, (255,255,255), 1)
                    #Mark the attendance of identified employee
                    attendance(name)
            #Display live feed
            cv2.imshow("Webcam Feed", image)
            key=cv2.waitKey(1) & 0xFF
            if key == ord('q'): 
                break      
        except:
            #Error exception 
            print("\n[-] Webcam not detected or the feed has been interrupted.")
            print("[-] Terminating!")
            break

    capture.release()
    cv2.destroyAllWindows()

if __name__ == "__main__":
    main()