How to erode this thresholded image using OpenCV

I am trying to first remove the captcha numbers by thresholding and then eroding it, to get slim continuous lines to get better output.

Problem: the eroded image is not continuous as you can see.

Original Image:

Threshold Image: (Here digits area is to thick, so I want it to be shrink, slim and continuous):

My output:

Desired Output:

code:

import os
import os.path
import cv2
import glob
import imutils
import matplotlib.pyplot as plt
import numpy as np
CAPTCHA_IMAGE_FOLDER = "generated_captcha_images"
OUTPUT_FOLDER = "extracted_letter_images"


# Get a list of all the captcha images we need to process
captcha_image_files = glob.glob(os.path.join(CAPTCHA_IMAGE_FOLDER, "*"))
counts = {}

# loop over the image paths
for (i, captcha_image_file) in enumerate(captcha_image_files):
    print("[INFO] processing image {}/{}".format(i + 1, len(captcha_image_files)))

    filename = os.path.basename(captcha_image_file)
    captcha_correct_text = os.path.splitext(filename)[0]

    image = cv2.imread(captcha_image_file)
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    thresh = cv2.threshold(
        gray, 36, 255, cv2.THRESH_BINARY_INV)[1]
    erode = cv2.erode(thresh, np.ones((2, 2), np.uint8), iterations=1)
    plt.imshow(erode, cmap="gray")
    plt.show()

In your case skeleton operation will work better

import cv2
import numpy as np

 
img = cv2.imread('/Users/alex/Downloads/fOrmgm.jpeg',0)

thinned = cv2.ximgproc.thinning(img)
element = cv2.getStructuringElement(cv2.MORPH_RECT, (3,3))
thinned = cv2.dilate(thinned, element) # this operation is optional

cv2.imshow("skeleton", thinned)
cv2.waitKey()

Update
You can find implementations of skeleton in python using morph operations. Result for me is not good. Probably you can filter this antennas.

Implementation of cv2.ximgproc.thinning() you can find here.