Detect beard from a face in Python

In this tutorial, we are going to learn how to detect beard form a face in Python using a webcam. We’ll be using Pycharm IDE to solve this problem. We will use OpenCV in order to recognize the face from video and also use Color Filtering to detect facial hair.

We will be using 3 windows to solve this problem:

Image: Displays the video captured by the webcam
Result: Only displays the required portion of the face i.e. nose, cheeks and chin region and the rest of the region is displayed as black
MASK: Complete video is black except for the portion where the beard is detected from the Result

Prerequisites

In order to build this program, we’ll require the following header files:

cv2
NumPy

We’ll also require a Haar Cascade in order to recognize a face from a given image/video. Here, we will be using haar cascade. Click on the highlighted link to download the XML file.

import cv2
import numpy as np
#Read XML file
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')

Here, we import the required files and read the date from the XML file.

Detect a face and the required region

Face Recognition from video in python using OpenCV

Then, we have to capture video from our webcam and detect a face by drawing a rectangle around the face region which is displayed by the Image Window.

After we detect the face, we only require the portion which includes the nose, cheeks and chin region which is to be displayed by the Result Window

For, this we draw an ellipse with white color around our required chin region on a black mask.

After this, we use a Bitwise operator in order to display the required region on the ellipse instead of just white color.

#create video capture object. 0 denotes webcam
cap=cv2.VideoCapture(0)

while True:
    # The input image to be recognised
    _, img=cap.read()

    #Convert the image to gray scale
    gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
    #Detect the faces
    faces = face_cascade.detectMultiScale(gray,1.1,4)
                                        #Gray Scale, Image factor, Number of Neighbours
#To construct a rectangle around the face detected
    for (x,y,w,h) in faces:


        # create a mask image of the same shape as input image, filled with 0s (black color)
        mask = np.zeros_like(img)
        # create a white filled ellipse
        mask = cv2.ellipse(mask, (int((x+w)/1.2), y+h),(69,69), 0, 0, -180, (255,255,255),thickness=-1)
        mask = cv2.cvtColor(mask, cv2.COLOR_BGR2RGB)



        # Bitwise AND operation to black out regions outside the mask
        result = np.bitwise_and(img, mask)

Detect a beard

Once we get the required region, we have to detect a beard. So in order to do this, we’ll try to find black color around this region.

Color filtering with OpenCV in Python

#Converting the final result as HSV inorder to detect colors
        hsv_img = cv2.cvtColor(result, cv2.COLOR_BGR2HSV)

        # Draws a rectangle

        cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2)

        # Black Color
        low_black = np.array([94, 80, 2])
        high_black = np.array([126, 255, 255])

        MASK = cv2.inRange(hsv_img, low_black, high_black)

Print Required Message

Now we want to print a message based on the output generated by MASK. For this, we will be using countNonZero() funciton from the cv2 library to check whether MASK is completely 0 i.e black or not.

#If the MASK only has black pixels caused due to no black colour/beard in the original image
if cv2.countNonZero(MASK) == 0:
    print("Beard Not Found")
else:
    print("Beard Found")

Display the Results

We use imshow() function from the cv2 library to show the video results.

Once the user clicks esc, the program terminates

#Display the Results
    cv2.imshow('Image',img)
    cv2.imshow('Result',result)
    cv2.imshow('MASK', MASK)
    #Wait for key press to close the image
    k=cv2.waitKey(30) &0xff
    #Break when esc key is pressed
    if k==27:
        break

cap.release()

Complete Code

Here is the complete code for this problem

import cv2
import numpy as np
#Read XML file
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')

#create video capture object. 0 denotes webcam
cap=cv2.VideoCapture(0)

while True:
    # The input image to be recognised
    _, img=cap.read()

    #Convert the image to gray scale
    gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
    #Detect the faces
    faces = face_cascade.detectMultiScale(gray,1.1,4)
                                        #Gray Scale, Image factor, Number of Neighbours
#To construct a rectangle around the face detected
    for (x,y,w,h) in faces:


        # create a mask image of the same shape as input image, filled with 0s (black color)
        mask = np.zeros_like(img)
        # create a white filled ellipse
        mask = cv2.ellipse(mask, (int((x+w)/1.2), y+h),(69,69), 0, 0, -180, (255,255,255),thickness=-1)
        mask = cv2.cvtColor(mask, cv2.COLOR_BGR2RGB)



        # Bitwise AND operation to black out regions outside the mask
        result = np.bitwise_and(img, mask)

        #Converting the final result as HSV inorder to detect colors
        hsv_img = cv2.cvtColor(result, cv2.COLOR_BGR2HSV)

        # Draws a rectangle

        cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2)

        # Black Color
        low_black = np.array([94, 80, 2])
        high_black = np.array([126, 255, 255])

        MASK = cv2.inRange(hsv_img, low_black, high_black)

        #If the MASK only has black pixels caused due to no black colour in the original image
        if cv2.countNonZero(MASK) == 0:
            print("Beard Not Found")
        else:
            print("Beard Found")



    #Display the Respective Results
    cv2.imshow('Image',img)
    cv2.imshow('Result',result)
    cv2.imshow('MASK', MASK)
    #Wait for key press to close the image
    k=cv2.waitKey(30) &0xff
    #Break when esc key is pressed
    if k==27:
        break

cap.release()

Result

No Beard Detected

Beard Detected