Unlock Python OpenCV: Load, Resize, Threshold, and Transform Images

1. Image Source

This image is sourced from Baidu Images; it is used solely for knowledge sharing.

2. Read and Display Image

imread(): read image

imshow(): display image

waitKey(): keep window open

import cv2
import numpy as np
# Read image
img = cv2.imread('girl.jpg')
# Show image
cv2.imshow('img', img)
# Wait indefinitely
cv2.waitKey(0)

Result:

3. Image Resizing

resize(): scale image; fx and fy are scaling factors (0.5 halves size)

import cv2
import numpy as np
img = cv2.imread('girl.jpg')
img = cv2.resize(img, None, fx=0.5, fy=0.5)
rows, cols, channels = img.shape
print(rows, cols, channels)
cv2.imshow('img', img)
cv2.waitKey(0)

Result:

4. Convert Image to HSV

Convert a color image to HSV to enable morphological operations.

import cv2
import numpy as np
img = cv2.imread('girl.jpg')
img = cv2.resize(img, None, fx=0.5, fy=0.5)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
cv2.imshow('hsv', hsv)
cv2.waitKey(0)

Result:

5. Binarize Image

Define lower and upper HSV bounds and apply inRange to obtain a binary mask.

import cv2
import numpy as np
img = cv2.imread('girl.jpg')
img = cv2.resize(img, None, fx=0.5, fy=0.5)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
lower_blue = np.array([90, 70, 70])
upper_blue = np.array([110, 255, 255])
mask = cv2.inRange(hsv, lower_blue, upper_blue)
cv2.imshow('mask', mask)
cv2.waitKey(0)

Result:

6. Erosion and Dilation

Apply erosion and dilation to reduce noise in the binary image.

import cv2
import numpy as np
img = cv2.imread('girl.jpg')
img = cv2.resize(img, None, fx=0.5, fy=0.5)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
lower_blue = np.array([90, 70, 70])
upper_blue = np.array([110, 255, 255])
mask = cv2.inRange(hsv, lower_blue, upper_blue)
erode = cv2.erode(mask, None, iterations=1)
dilate = cv2.dilate(erode, None, iterations=1)
cv2.imshow('erode', erode)
cv2.imshow('dilate', dilate)
cv2.waitKey(0)

Result:

7. Replace Pixels by Color

Iterate over each pixel of the eroded mask; replace white pixels with red in the original image.

import cv2
import numpy as np
img = cv2.imread('girl.jpg')
img = cv2.resize(img, None, fx=0.5, fy=0.5)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
lower_blue = np.array([90, 70, 70])
upper_blue = np.array([110, 255, 255])
mask = cv2.inRange(hsv, lower_blue, upper_blue)
erode = cv2.erode(mask, None, iterations=1)
rows, cols, _ = img.shape
for i in range(rows):
    for j in range(cols):
        if erode[i, j] == 255:
            img[i, j] = (0, 0, 255)  # BGR red
cv2.imshow('result', img)
cv2.waitKey(0)

Result: