从具有轮廓的图像中提取 ROI

Question

I have bunch of pictures like this :

And want to extract this part :

so first i use masking the red part, secondly i threshold and contour. But still cant select the only lip part.

import cv2 
import matplotlib.pyplot as plt
import numpy as np
img = cv2.imread("C:/Users/TBD/Desktop/TEZ GORSEL/098.JPG")

img_hsv=cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

lower_red = np.array([0,95,50])
upper_red = np.array([2,255,255])
mask0 = cv2.inRange(img_hsv, lower_red, upper_red)

lower_red = np.array([175,95,50])
upper_red = np.array([180,255,255])
mask1 = cv2.inRange(img_hsv, lower_red, upper_red)

mask = mask0+mask1
output_img = img.copy()
output_img[np.where(mask==0)] = 0

output_hsv = img_hsv.copy()
output_hsv[np.where(mask==0)] = 0
son = cv2.cvtColor(output_hsv, cv2.COLOR_HSV2BGR)
gray_son = cv2.cvtColor(son, cv2.COLOR_BGR2GRAY)
ret, thresh = cv2.threshold(gray_son, 10, 255, cv2.THRESH_BINARY)
contours,hierarchy = cv2.findContours(thresh, 1, 2)
for cnt in contours:
    x,y,w,h = cv2.boundingRect(cnt)
    if w>1960 and h>800:
        break
        
cropped = son[y:y+h, x:x+w]
cv2.imwrite('cropped.jpg', cropped)

after these code i can reach this

But still i cannot extract the lip only. What should i do next?

Answer 1

With classical image processing as your solution, this will change the results if the environmental conditions change. Using face landmarks you can get detailed information and crop it from the image as a rectangle.

Other solution can be that firstly, you can detect mouth using in below afterwards, you can apply your solution method or lip detect.

import cv2
import numpy as np

mouth_cascade = cv2.CascadeClassifier('./cascade_files/haarcascade_mcs_mouth.xml')

if mouth_cascade.empty():
  raise IOError('Unable to load the mouth cascade classifier xml file')

cap = cv2.VideoCapture(0)
ds_factor = 0.5

while True:
    ret, frame = cap.read()
    frame = cv2.resize(frame, None, fx=ds_factor, fy=ds_factor, interpolation=cv2.INTER_AREA)
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    mouth_rects = mouth_cascade.detectMultiScale(gray, 1.7, 11)
    for (x,y,w,h) in mouth_rects:
        y = int(y - 0.15*h)
        cv2.rectangle(frame, (x,y), (x+w,y+h), (0,255,0), 3)
        break

    cv2.imshow('Mouth Detector', frame)

    c = cv2.waitKey(1)
    if c == 27:
        break

cap.release()
cv2.destroyAllWindows()