OpenCV: Crop an image based on certain condition

Question

I need to crop the image when ever there's a blank space detected.

and the expected output is the individual cropped images of the input data like below after blank space is detected between two text in an image.

Any idea would be appreciated!

Answer 1

Here is how I suggest to do that in Python/Opencv from my comment above.

"Threshold on the green using cv2.inRange(). Use a horizontal morphology close such that the letters space is closed, but not so large that the space between words is filled. Then get contours and get the bounding boxes. Then use the bounding boxes to crop the original image."

Input:

import cv2
import numpy as np

# read the image
img = cv2.imread('StandOutFrom.png')

# threshold on green
lower=(30,30,0)
upper=(110,110,60)
thresh = cv2.inRange(img, lower, upper)

# apply horizontal close morphology
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (15,3))
morph = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, kernel)

# get external contours
contours = cv2.findContours(morph, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
contours = contours[0] if len(contours) == 2 else contours[1]

# get bounding boxes and save to disk
i = 1
result = img.copy()
for cntr in contours:
    # get bounding boxes
    x,y,w,h = cv2.boundingRect(cntr)
    cv2.rectangle(result, (x, y), (x+w, y+h), (0, 0, 255), 2)
    print(i,x,y,w,h)
    crop = img[y:y+h, x:x+w]
    cv2.imwrite("StandOutFrom_crop_{0}.png".format(i), crop)
    i = i + 1

# save result
cv2.imwrite('StandOutFrom_thresh.png', thresh )
cv2.imwrite('StandOutFrom_morph.png', morph )
cv2.imwrite('StandOutFrom_bounding_boxes.png', result )

# show results
cv2.imshow('thresh', thresh)
cv2.imshow('morph', morph)
cv2.imshow('boundingboxes', result)
cv2.waitKey(0)

Threshold Image:

Morphology Close Image:

Bounding Box Image:

Result Images: