Find number of close loops in Binary Edge Image

问题

I have a binary image with the number of close loops and free curves like shown in image..

My approach so far- (python code):

Skeletonize the image to get edge of 1 pixel
Use label/bwlabel to get individual curve/edge
Applied DFS to get the edge which is not close
Apply DFS to get close edge..but not working correctly..

Expected output -

Number of close loops =6, the sum of the pixels along the loop or all (x,y) points along the edge of the close curve as highlighted in RED

Expected Output-

回答1:

You already have a binary image: simply use findContours() to get the inner contours.

Bare in mind you want use cv2.RETR_CCOMP as the Countour Retrieval Mode to easily access the holes:

retrieves all of the contours and organizes them into a two-level hierarchy. At the top level, there are external boundaries of the components. At the second level, there are boundaries of the holes. If there is another contour inside a hole of a connected component, it is still put at the top level.

It should be a matter of simply getting the length of countours on the second level (holes)

回答2:

Here is one way to do that in Python/OpenCV.

Read the input as grayscale
Threshold to binary
Get the contours and hierarchy
Loop over the contours and corresponding hierarchy and discard contours with no parent (the external ones) and save and count the contours with no children (the innermost ones)
Draw the contours on the input
Save the results

Input:

import cv2
import numpy as np

# read image as grayscale
img = cv2.imread('loops.png', cv2.IMREAD_GRAYSCALE)
hh, ww = img.shape

# blacken right columns that are white
img[0:hh, ww-3:ww] = 0

# threshold
thresh = cv2.threshold(img, 0, 255, cv2.THRESH_BINARY)[1]

# get contours
contours = cv2.findContours(thresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
hierarchy = contours[1] if len(contours) == 2 else contours[2]
contours = contours[0] if len(contours) == 2 else contours[1]

# get the actual inner list of hierarchy descriptions
hierarchy = hierarchy[0]

# count inner contours
count = 0
result = img.copy()
result = cv2.merge([result,result,result])
for component in zip(contours, hierarchy):
    cntr = component[0]
    hier = component[1]
    # discard outermost no parent contours and keep innermost no child contours
    # hier = indices for next, previous, child, parent
    # no parent or no child indicated by negative values
    if (hier[3] > -1) & (hier[2] < 0):
        count = count + 1
        cv2.drawContours(result, [cntr], 0, (0,0,255), 2)

# print count
print("count:",count)

# save result
cv2.imwrite("loops_result.png",result)

# show result   
cv2.imshow("result", result)
cv2.waitKey(0)
cv2.destroyAllWindows()

Resulting contours:

Count:

count: 6

来源：https://stackoverflow.com/questions/62120690/find-number-of-close-loops-in-binary-edge-image

标签

python

image

OpenCV

hough-transform