OpenCV findContours() detects contours only if the image is saved and read beforehand

Providing some context:
I am trying to get the number of boxes in this image,

I have the above image stored in an ndarray blank_img.

If I run the following:

v = np.median(blank_img)
sigma = 0.33
lower = int(max(0, (1.0 - sigma) * v))
upper = int(min(255, (1.0 + sigma) * v))
edges = cv2.Canny(blank_img, lower, upper, 3)
edges = cv2.dilate(edges, np.ones((2, 2), dtype=np.uint8))

cnts= cv2.findContours(edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]

len(cnts) is equal to 1.
But if i save the array blank_img before the canny edge detection and the findContours, as follows:

cv2.imwrite('boxes.jpg', blank_img)

blank_img = cv2.imread('boxes.jpg')

Now when i run my findContour code snippet, len(cnts) is > 1. I would like to know how to fix it so that I don’t need to have an image saving overhead.

~~EDIT~~
Find the following code which I used to create blank_img

blank_img = 255*np.zeros((height, width, 3), np.uint8)
for line in lines:
    x1, y1, x2, y2 = line
    cv2.line(blank_img,(x1,y1),(x2,y2),(255,255,255),1)

Where lines is a list of lines returned by HoughLinesP:

lines = cv2.HoughLinesP(edges, 1, np.pi / 180, 100, None,minlinelength,maxlinegap)