How to apply erosion for only thick lines in images - python-imaging-library

Question

Need to apply erosion only to the lines in my image which are more thicker using python
Here is my input image
I'm having a image which contains white lines with both thick and thin lines,my goal is to erode only the lines which are more thicker using python.i used normal erosion using opencv when applying this method thin lines are removing from the image.Erosion has to apply only on thick lines
import cv2
import numpy as np
img = cv2.imread('123.png',0)
kernel = np.ones((6,6),np.uint8)
erosion = cv2.erode(img,kernel,iterations = 1)
cv2.imshow("Result", erosion)
cv2.waitKey(0)
cv2.destroyAllWindows()
How can i achieve this any answers will be highly useful for me.Thanks in advance

Answer 1

Here is one way to do that in Python/OpenCV.
Read the input as grayscale
Threshold it
Apply morphology close to remove the thin lines and save as a mask image
Apply morphology dilate to the threshold image again to thin the lines
Use np.where() to combine the threshold and dilated images using the mask
Save the results
Input:
import cv2
import numpy as np
# read the input as grayscale
img = cv2.imread('black_lines.jpg', cv2.IMREAD_GRAYSCALE)
# threshold
thresh = cv2.threshold(img, 0, 255, cv2.THRESH_BINARY+cv2.THRESH_OTSU)[1]
# apply morphology close to remove the thin lines and save as mask
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5,5))
mask = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, kernel)
# apply morphology dilate to thin image as desired
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5,5))
dilate = cv2.morphologyEx(thresh, cv2.MORPH_DILATE, kernel)
# combine the thresh and dilate using mask
result = np.where(mask==0, dilate, thresh)
# save results
cv2.imwrite('black_lines_mask.jpg', mask)
cv2.imwrite('black_lines_dilate.jpg', dilate)
cv2.imwrite('black_lines_thinned.jpg', result)
# show results
cv2.imshow('mask', mask)
cv2.imshow('dilate', dilate)
cv2.imshow('thinned', result)
cv2.waitKey(0)
Mask Image:
Dilate Image:
Combined Image with thin lines: