PyFry.py

import cv2
import os
import dlib
import tkinter as tk
from PIL import Image, ImageOps, ImageEnhance
from utils.utils import Colors
from imutils import face_utils
from tkinter import filedialog

'''
TODO: -> Compressing (Crushing) and back (to increase noise) :: DONE
      -> Applying Red and Orange hue filters for classic deep fry look :: DONE
      -> Detecting eye coordinates and applying the deepfry eye flare in the center::DONE

'''

def userInput():
up-and-structure-issue12
    '''
    Allowing user to choose the image that has to be deepfried
    '''
    root =tk.Tk()
    root.withdraw()
    global filepath
    filepath = list(root.tk.splitlist(filedialog.askopenfilenames(title="PyFry - Choose Image")))
    print("picture location = ",filepath)

    
def irisCoords(eye):
    '''
    Finding the center point of the eye using the average outer extremes average of the eyes
    '''
    mid = (eye[0] +eye[3])/2
    mid = (int(mid[0]), int(mid[1]))
    return mid


def generateHue(img):
    '''
    Generating and increasing prominency of red band of the image
    '''
    img = img.convert('RGB')
    red = img.split()[0] #(R,G,B)
    red = ImageEnhance.Contrast(red).enhance(2.0)
    red = ImageEnhance.Brightness(red).enhance(1.5)
    red = ImageOps.colorize(red, Colors.RED, Colors.YELLOW)
    img = Image.blend(img, red, 0.77)
    #Keeping a 100% sharpness value for now, But would probably be better with a higher sharpness value
    img = ImageEnhance.Sharpness(img).enhance(150)
    return img


def crushAndBack(img):
    img = img.convert('RGB')
    w,h = img.width, img.height
    img = img.resize((int(w ** .95), int(h ** .95)), resample=Image.LANCZOS)
    img = img.resize((int(w ** .90), int(h ** .90)), resample = Image.BILINEAR)
    img = img.resize((int(w ** .90), int(h ** .90)), resample = Image.BICUBIC)
    img = img.resize((w,h), resample = Image.BICUBIC)
    return img


def addFlare(img):
    ''' 
    Initialising dlib for frontal facial features 
    '''
    
    flare = Image.open('flare.png')
    detect = dlib.get_frontal_face_detector()
    predict = dlib.shape_predictor("assets\shape_predictor_68_face_landmarks.dat")

    (lS, lE) = face_utils.FACIAL_LANDMARKS_68_IDXS["left_eye"]
    (rS, rE) = face_utils.FACIAL_LANDMARKS_68_IDXS["right_eye"]
    
    imgCV = cv2.imread('test.jpg')
    #imgCV = cv2.imread('test2.jpg')

    gray = cv2.cvtColor(imgCV, cv2.COLOR_BGR2GRAY)
    subjects = detect(gray, 0)

    for subject in subjects:
        shape = predict(gray, subject)
        shape = face_utils.shape_to_np(shape)
        leftEye = shape[lS:lE]
        rightEye = shape[rS:rE]

    '''
        Assigning an area to paste the flare png Using the coordinates given by the Dlib module
        ln,rn is the distance between the top left and bottom right of the iris multiplied by 4.
        This is used to find the basic coordinates of the area in which the flare image will be pasted
    '''

    rn=(rightEye[4][0]-rightEye[0][0])*3
    ln=(leftEye[4][0]-leftEye[0][0])*3

    rec0=(leftEye[1][0]-ln,leftEye[1][1]-ln)
    rec1=(leftEye[4][0]+ln,leftEye[4][1]+ln)
     
    rec2=(rightEye[1][0]-rn,rightEye[1][1]-rn)
    rec3=(rightEye[4][0]+rn,rightEye[4][1]+rn)
    
    print("Area for left eye",rec0,rec1)
    print("Area for right eye",rec2,rec3)

    #Area Assignment for left eye and right eye
    areaLeft=(rec0[0],rec0[1],rec1[0],rec1[1])
    areaRight=(rec2[0],rec2[1],rec3[0],rec3[1])
    
    #Resizing the flare image to fit the area
    flareLeft=flare.resize((rec1[0]-rec0[0],rec1[1]-rec0[1]))
    flareRight=flare.resize((rec3[0]-rec2[0],rec3[1]-rec2[1]))
    
    #Pasting the flare image on the area.
    #Third parameter is an alpha channel that provides transparency for the png
    img.paste(flareLeft,areaLeft,flareLeft)
    img.paste(flareRight,areaRight,flareRight)
    return img


def main():
    userInput()
    
    for img_path in filepath:
        img = Image.open(img_path)
        #img = Image.opne('test2.jpg')
        img = img.convert('RGB')
        img = crushAndBack(img)
        img = generateHue(img)
        img = addFlare(img)
        
        img.show()
        #img.save('output2.jpg')
        filename = os.path.splitext(os.path.basename(img_path))[0]
        img.save('%s_output.jpg' % filename)
        print("output saved as %s_output.jpg" % filename)
   

if __name__ == '__main__':
    main()