Converting an RGBA image to Grayscale and Binary

[Joni_Engr]

Hi, I have a RGBA image which I need to convert to Grayscale and then to binary with a threshold to get binary histogram but I have 2D Grayscale image in my code which I don't understand how to convert to binary.

# Import libraries 
from PIL import Image
import matplotlib.pyplot as plt
import numpy as np
import cv2

# Reading an Image
image = Image.open('Board1_1.png')

# Properties of an Image
print(image.size)
print(image.format)
print(image.mode)

# Grayscale conversion 
image_gray = image.convert('LA')

image_gray_array = 255 - np.asarray(image_gray)

image_gray_array_1 = image_gray_array[:,:,0]
image_gray_array_2 = image_gray_array[:,:,1]

image_gray.show()

[Pedroski55]

I believe Python's module cv2 can do just about anything with images, and a lot more besides!

This code below gives me a threshold image. It's an excerpt from my own little amateurish OMR programme.

I save the steps on the way just so I can see what is going on.

This is all a bit on the "threshold" of my understanding! But it works!

import cv2

destination = '/home/pedro/omr_steps/'

f = jpg_files[0]
original = cv2.imread(f)
name = 'step0_original.png'
cv2.imwrite(destination + name, original)

# crop the bottom part of the image, the part with the choices, the part to mark
# f is just a file path: savename = savepathMarked + 'temp_cropped_image.png'
cropped_image = cropAll(f, destination, tup)
# im is the image to write the stroked contours on
im = cv2.imread(cropped_image)
name = 'step1_cropped_image.png'
cv2.imwrite(destination + name, im)

# make a grey image
grey = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)
name = 'step2_grey.png'
cv2.imwrite(destination + name, grey)

# blur the grey image this helps the next step: Canny
blurred = cv2.GaussianBlur(grey, (5, 5), 0)
name = 'step3_blurred.png'
cv2.imwrite(destination + name, blurred)

# get the edges of the blurred image
edged = cv2.Canny(blurred, 75, 200)
name = 'step4_edged.png'
cv2.imwrite(destination + name, edged)

# find contours in the edge map, then use the contours on the thresh
    
edged2 = edged.copy()
contours, hierarchy = cv2.findContours(edged2, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
print('reading edged2, there are ', len(contours), ' contours')

thresh = cv2.threshold(grey, 0, 255, cv2.THRESH_BINARY_INV | cv2.THRESH_OTSU)[1]
name = 'step5_thresh.png'
cv2.imwrite(destination + name, thresh)

[Joni_Engr]

Hi, I get an error "ModuleNotFoundError: No module named 'cv2'"

I have already installed "pip install opencv-python" but still get an error

[Pedroski55]

In your post you had it imported! Why, if you are not using it?

I only use Linux, can't help if you use Windoze!

Normally, I install a module with pip, then it is available.