使用opencv python从表单中检测复选框

Question

给定一个牙科表格作为输入，需要使用图像处理找到表格中存在的所有复选框。我在下面回答了我目前的方法。有没有更好的方法来查找低质量文档的复选框？

样本输入：

Answer 1

这是我们可以解决问题的一种方法，

import cv2
import numpy as np
image=cv2.imread('path/to/image.jpg')

### binarising image
gray_scale=cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
th1,img_bin = cv2.threshold(gray_scale,150,225,cv2.THRESH_BINARY)

定义垂直和水平内核

lineWidth = 7
lineMinWidth = 55
kernal1 = np.ones((lineWidth,lineWidth), np.uint8)
kernal1h = np.ones((1,lineWidth), np.uint8)
kernal1v = np.ones((lineWidth,1), np.uint8)

kernal6 = np.ones((lineMinWidth,lineMinWidth), np.uint8)
kernal6h = np.ones((1,lineMinWidth), np.uint8)
kernal6v = np.ones((lineMinWidth,1), np.uint8)

检测水平线

img_bin_h = cv2.morphologyEx(~img_bin, cv2.MORPH_CLOSE, kernal1h) # bridge small gap in horizonntal lines
img_bin_h = cv2.morphologyEx(img_bin_h, cv2.MORPH_OPEN, kernal6h) # kep ony horiz lines by eroding everything else in hor direction

寻找垂直线

## detect vert lines
img_bin_v = cv2.morphologyEx(~img_bin, cv2.MORPH_CLOSE, kernal1v)  # bridge small gap in vert lines
img_bin_v = cv2.morphologyEx(img_bin_v, cv2.MORPH_OPEN, kernal6v)# kep ony vert lines by eroding everything else in vert direction

合并垂直和水平线以获得块。添加一层扩张以消除小间隙

### function to fix image as binary
def fix(img):
    img[img>127]=255
    img[img<127]=0
    return img

img_bin_final = fix(fix(img_bin_h)|fix(img_bin_v))

finalKernel = np.ones((5,5), np.uint8)
img_bin_final=cv2.dilate(img_bin_final,finalKernel,iterations=1)

对二值图像应用连通分量分析以获得所需的块。

ret, labels, stats,centroids = cv2.connectedComponentsWithStats(~img_bin_final, connectivity=8, ltype=cv2.CV_32S)

### skipping first two stats as background
for x,y,w,h,area in stats[2:]:
    cv2.rectangle(image,(x,y),(x+w,y+h),(0,255,0),2)