Smile Detection using Opencv-Python

Smile Detection using Opencv-Python

Welcome back to a another Blog on Opencv.

In this Blog we will recognize Smile on face using an Image.

so lets looks at requirements.

Requirements.

1. Opencv (pip install opencv-python)
2. numpy   (automatically install with opencv)
3. python 3.x

So lets firsts code and then Understand its working.

       			 
# Dynamic Coding
# Smile detection using Opencv-Python
# Import library
import cv2

#lets  include the haar cascade xml file.
detection = cv2.CascadeClassifier('haarcascade_smile.xml');

#reading the image
image = cv2.imread('tom.png') 
#converting to gray scale
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)


faces = detection.detectMultiScale(gray, 1.3, 20)
#loop for drawing rectangle
for(x, y, w, h) in faces: 
    #drawing rectangle
    cv2.rectangle(image, (x,y), (x+w, y+h), (0,255,0), 2)  
    
#displaying image in new window
cv2.imshow('WINDOW', image)

#Saving image to folder
#cv2.imwrite('tom_modified.jpg', image)
cv2.waitKey()

 

Output.

How it Works ?

In the first 2 line we first import the module and make a instance for cascade classifier class which takes haarcascade_smile.xml as parameter.

       			 
import cv2
#lets  include the haar cascade xml file.
detection = cv2.CascadeClassifier('haarcascade_smile.xml');
 

the haarcascade_smile.xml is the core file for our whole code because it is used to recognize the smile on image.

Dowload  haarcascade_smile.xml from here .

Now read the Image using imread method of opencv and then convert the colored image into gray scale image for better results.

       			 
image = cv2.imread('tom.png') 
#converting to gray scale
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
 

After reading and converting image into gray scale used the cascade classifier object with Detectmultiscale method which detects the smile according to the data present in the haarcascade_smile.xml file and it returns 4 values which are the coordinates of the image and we use those 4 returned values for creating a rectangle over the image , means over the smile.

       			 
faces = detection.detectMultiScale(gray, 1.3, 20)
#loop for drawing rectangle
for(x, y, w, h) in faces: 
    #drawing rectangle
    cv2.rectangle(image, (x,y), (x+w, y+h), (0,255,0), 2)
 

now all the work is done so lets see our image using imshow method and waitkey method for holding the  image screen on the Window.

       			 
cv2.imshow('WINDOW', image)
#Saving image to folder
#cv2.imwrite('tom_modified.jpg', image)
cv2.waitKey()
 

 you can also use imwrite method to save the smile recognized image.

---

I hope you like the Blog and learn some thing new from it .

if you have any doubts and errors please comments we will reply you as soon as possible.