Box Blur Algorithm - With Python implementation
Last Updated :
30 Dec, 2020
Improve
The pixels in an image are represented as integers. After blurring each pixel 'x' of the resulting image has a value equal to the average of the pixels surrounding 'x' including 'x'. For example, consider a 3 * 3 image as
image = \begin{bmatrix}
1 & 1 & 1\\
1 & 7 & 1\\
1 & 1 & 1
\end{bmatrix}
Then, the resulting image after blur is blurred_image =
\begin{bmatrix}
1
\end{bmatrix}
So, the pixel of blurred image is calculated as (1 + 1 + 1 + 1 + 7 + 1 + 1 + 1 + 1) / 9 = 1.66666 = 1
image = \begin{bmatrix}
7 & 4 & 0 & 1\\
5 & 6 & 2 & 2\\
6 & 10 & 7 & 8\\
1 & 4 & 2 & 0
\end{bmatrix}
It's blurred image is given below:
blurred\_image = \begin{bmatrix}
5 & 4\\
4 & 4
\end{bmatrix}
Explanation:
There are four 3 * 3 matrix possible in the above image. So there are 4 blurred pixel in the resulting image. The four matrices are:
Python3 1==
Output:
Python3 1==
Output:
Box Blur Algorithm -
Box blur is also known as box linear filter. Box blurs are frequently used to approximate Gaussian blur. A box blur is generally implemented as an image effect that affects the whole screen. The blurred colour of the current pixel is the average of the current pixel's colour and its 8 neighbouring pixels.Note: For each 3 * 3 pixel matrix there is one blurred pixel calculated as above.FOr Example, Consider the below image.
Implementation in Python:\begin{bmatrix} 7 & 4 & 0\\ 5 & 6 & 2\\ 6 & 10 & 7 \end{bmatrix} ,\begin{bmatrix} 4 & 0 & 1\\ 6 & 2 & 2\\ 10 & 7 & 8 \end{bmatrix} ,\begin{bmatrix} 5 & 6 & 2\\ 6 & 10 & 7\\ 1 & 4 & 2 \end{bmatrix} , and\begin{bmatrix} 6 & 2 & 2\\ 10 & 7 & 8\\ 4 & 2 & 0 \end{bmatrix}
def square_matrix(square):
""" This function will calculate the value x
(i.e. blurred pixel value) for each 3 * 3 blur image.
"""
tot_sum = 0
# Calculate sum of all the pixels in 3 * 3 matrix
for i in range(3):
for j in range(3):
tot_sum += square[i][j]
return tot_sum // 9 # return the average of the sum of pixels
def boxBlur(image):
"""
This function will calculate the blurred
image for given n * n image.
"""
square = [] # This will store the 3 * 3 matrix
# which will be used to find its blurred pixel
square_row = [] # This will store one row of a 3 * 3 matrix and
# will be appended in square
blur_row = [] # Here we will store the resulting blurred
# pixels possible in one row
# and will append this in the blur_img
blur_img = [] # This is the resulting blurred image
# number of rows in the given image
n_rows = len(image)
# number of columns in the given image
n_col = len(image[0])
# rp is row pointer and cp is column pointer
rp, cp = 0, 0
# This while loop will be used to
# calculate all the blurred pixel in the first row
while rp <= n_rows - 3:
while cp <= n_col-3:
for i in range(rp, rp + 3):
for j in range(cp, cp + 3):
# append all the pixels in a row of 3 * 3 matrix
square_row.append(image[i][j])
# append the row in the square i.e. 3 * 3 matrix
square.append(square_row)
square_row = []
# calculate the blurred pixel for given 3 * 3 matrix
# i.e. square and append it in blur_row
blur_row.append(square_matrix(square))
square = []
# increase the column pointer
cp = cp + 1
# append the blur_row in blur_image
blur_img.append(blur_row)
blur_row = []
rp = rp + 1 # increase row pointer
cp = 0 # start column pointer from 0 again
# Return the resulting pixel matrix
return blur_img
# Driver code
image = [[7, 4, 0, 1],
[5, 6, 2, 2],
[6, 10, 7, 8],
[1, 4, 2, 0]]
print(boxBlur(image))
[[5, 4], [4, 4]]Test case 2:
image = [[36, 0, 18, 9],
[27, 54, 9, 0],
[81, 63, 72, 45]]
print(boxBlur(image))
[[40, 30]]Further Read: Box Blur using PIL library | Python
