One-Click Background Replacement for ID Photos Using Python OpenCV

In daily life we often need ID photos with different background colors (red, white, blue). This guide shows how to replace the background of an ID photo automatically without manual cropping, using Python and OpenCV.

Knowledge Points

1. Image processing 2. OpenCV 3. NumPy 4. Basic Python knowledge

Environment

Windows, PyCharm, Python 3

Steps

1. Import libraries

import numpy as np
import cv2

2. Install OpenCV pip install OpenCV-python 3. Load image img = cv2.imread('timg.jpg') 4. Resize image

rows, cols, channels = img.shape
print(rows, cols, channels)
img = cv2.resize(img, None, fx=0.5, fy=0.5)
rows, cols, channels = img.shape
print(rows, cols, channels)

5. Display and convert to HSV, define color range

cv2.imshow('img', img)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
cv2.imshow('hsv', hsv)
lower_blue = np.array([90, 70, 90])
upper_blue = np.array([110, 255, 255])

6. Create binary mask

mask = cv2.inRange(hsv, lower_blue, upper_blue)
cv2.imshow('Mask', mask)

7. Erosion

erosion = cv2.erode(mask, None, iterations=1)
cv2.imshow('erosion', erosion)

8. Dilation

dilation = cv2.dilate(mask, None, iterations=1)
cv2.imshow('dilation', dilation)

9. Replace background color where mask is white

for i in range(rows):
    for j in range(cols):
        if dilation[i, j] == 255:
            img[i, j] = (0, 0, 255)  # BGR (red)
cv2.imshow('res', img)

10. Save result and clean up

cv2.imwrite('ting.png', img)
cv2.waitKey(0)  # wait indefinitely
cv2.destroyAllWindows()  # close all windows

The script processes the input image, isolates the desired background color, replaces it with red (or any chosen color), and saves the final image, providing a quick solution for generating ID photos with custom backgrounds.