*本文介绍图像形态学的应用和使用图像形态学进行骨架提取算法。核心心代码均来自互联网。*

一、提取水平垂直线

示例1 五线谱

#! /usr/bin/env python
# -*- coding:utf-8 -*-
import cv2

src = cv2.imread("1.png")
gray_src = cv2.cvtColor(src, cv2.COLOR_BGR2GRAY)
cv2.imshow("input image", src)
cv2.imshow("gray image", gray_src)
gray_src = cv2.bitwise_not(gray_src)
binary_src = cv2.adaptiveThreshold(gray_src, 255, cv2.ADAPTIVE_THRESH_MEAN_C, cv2.THRESH_BINARY, 15, -2)
cv2.imshow("binary_src", binary_src)
# 提取水平线
#hline = cv2.getStructuringElement(cv2.MORPH_RECT, ((src.shape[1] // 16), 1), (-1, -1))
# 提取垂直线
vline = cv2.getStructuringElement(cv2.MORPH_RECT, (1, (src.shape[0] // 16)), (-1, -1))
# 形态学的开操作——先腐蚀再膨胀，合并用morphologyEx
# temp = cv2.erode(binary_src, hline)
# dst = cv2.dilate(temp, hline)
dst = cv2.morphologyEx(binary_src, cv2.MORPH_OPEN, vline)
dst = cv2.bitwise_not(dst)
cv2.imshow("Final image", dst)
cv2.waitKey(0)

示例2 三线格

#! /usr/bin/env python
# -*- coding:utf-8 -*-
import cv2

src = cv2.imread("1.png")
gray_src = cv2.cvtColor(src, cv2.COLOR_BGR2GRAY)
cv2.imshow("input image", src)
cv2.imshow("gray image", gray_src)
gray_src = cv2.bitwise_not(gray_src)
binary_src = cv2.adaptiveThreshold(gray_src, 255, cv2.ADAPTIVE_THRESH_MEAN_C, cv2.THRESH_BINARY, 15, -2)
cv2.imshow("binary_src", binary_src)
# 提取垂直线
vline = cv2.getStructuringElement(cv2.MORPH_RECT, (1, (src.shape[0] // 29)), (-1, -1))
# 形态学的开操作——先腐蚀再膨胀，合并用morphologyEx
dst = cv2.morphologyEx(binary_src, cv2.MORPH_OPEN, vline)
dst = cv2.bitwise_not(dst)
cv2.imshow("Final image", dst)
cv2.waitKey(0)

二、形态学骨架提取

图像骨架提取是提取目标在图像上的中心像素轮廓。关于骨架提取，现存的算法有一千种以上，分为迭代和非迭代两大类。在迭代算法中，又分为并行迭代和顺序迭代两种。使用腐蚀和膨胀操作就可以获得图像骨架。

#! /usr/bin/env python
# -*- coding:utf-8 -*-
import numpy as np
import cv2
import sys

im_path = 'hello.png'
im = cv2.imread(im_path, 0)

if im is None:
 sys.exit()

ret, im = cv2.threshold(im, 127, 255, cv2.THRESH_BINARY)
element = cv2.getStructuringElement(cv2.MORPH_CROSS, (3, 3))

skel = np.zeros(im.shape, np.uint8)
erode = np.zeros(im.shape, np.uint8)
temp = np.zeros(im.shape, np.uint8)

i = 0
while True:
 cv2.imshow('im %d' % i, im)
 erode = cv2.erode(im, element)
 temp = cv2.dilate(erode, element)

 # 消失的像素是skeleton的一部分
 temp = cv2.subtract(im, temp)
 skel = cv2.bitwise_or(skel, temp)
 im = erode.copy()

 if cv2.countNonZero(im) == 0:
 break;
 i += 1

cv2.imshow('Skeleton', skel)
cv2.waitKey()