Catalog

1 What is image contour

2 Find the outline

3 Draw the outline

4 The area and perimeter of the contour

5 Polygon approximation and convex hull

6. convex hull

 7.  Outside rectangle

1 What is image contour

An image outline is a curve of continuous points with the same color or gray scale . Contour is very useful in shape analysis and object detection and recognition .

Function of contour :

• For graphic analysis

• Object recognition and detection

Be careful :

• In order to test the accuracy , You need to edit the image first Two valued or Canny operation .

• When drawing the outline, the input image will be modified , If you want to continue using the original image later , The original image should be stored in other variables .

2 Find the outline

findContours(image, mode, method[, contours[, hierarchy[, offset]]])

• mode Find the pattern of the contour

  • RETR_EXTERNAL = 0, Indicates that only the peripheral contour is detected

  • RETR_LIST = 1, The detected contour does not establish a hierarchical relationship , That is, detect all contours , 

  • RETR_CCOMP = 2, At most two levels per floor , From small to large , From inside to outside .

  • RETR_TREE = 3, Store the outline according to the tree , From big to small , From right to left . （ Commonly used ）

• method The contour approximation method is also called ApproximationMode

  • CHAIN_APPROX_NONE Save points on all contours

  • CHAIN_APPROX_SIMPLE, Save only corners , Like a quadrilateral , Keep only quadrilateral 4 Corner , Less storage information , Compare Commonly used

• return contours and hierachy That is, outline and level （ You can also use the return value to receive ）

3 Draw the outline

drawContours(image, contours, contourIdx, color[, thickness[, lineType[, hierarchy[, maxLevel[, offset]]]]])

• image Outline image to be drawn

• contours Outline point （ from findContours Data found ）

• contourIdx The number of the outline to be drawn . -1 Means to draw all the contours

• color Color of outline , Such as (0, 0, 255) It means red

• thickness Line width , -1 Means to fill all

Code case ：

import cv2
import numpy as np

#  The image display effect is black and white ,  But it's actually 3 Color images of channels .
img = cv2.imread('./contours1.jpeg')

#  Become a single channel black-and-white picture 
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

#  Two valued ,  Pay attention to it 2 Return values ,  Thresholds and results 
ret, binary = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)


#  Contour search ,  The new version returns two results ,  Outline and hierarchy ,  The old version returns 3 Parameters ,  Images ,  Outline and hierarchy 
result, contours, hierarchy = cv2.findContours(binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

#  Draw the outline ,  Be careful ,  Drawing the outline will change the original drawing 
cv2.drawContours(img, contours, 1, (0, 0, 255), 2)

cv2.imshow('img', img)

cv2.waitKey(0)
cv2.destroyAllWindows()

4 The area and perimeter of the contour

Contour area refers to the area surrounded by all pixels in each contour , In pixels .

Contour area is one of the important statistical characteristics of contour , The implicit information of each contour can be further analyzed by the size of the contour area , For example, distinguish the size of objects by contour area and recognize different objects .

After finding the contour , There may be many small outlines , We can filter by the area of the contour .

• contourArea(contour)

• contour Contour data

• arcLength(curve, closed)

  • curve That is, contour

  • closed Whether it is a closed contour

Case code ：

import cv2
import numpy as np

#  The image display effect is black and white ,  But it's actually 3 Color images of channels .
img = cv2.imread('./contours1.jpeg')

#  Become a single channel black-and-white picture 
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

#  Two valued ,  Pay attention to it 2 Return values ,  Thresholds and results 
ret, binary = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)


#  Contour search ,  The new version returns two results ,  Outline and hierarchy ,  The old version returns 3 Parameters ,  Images ,  Outline and hierarchy 
result, contours, hierarchy = cv2.findContours(binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

#  Draw the outline ,  Be careful ,  Drawing the outline will change the original drawing 
cv2.drawContours(img, contours, 1, (0, 0, 255), 2)

#  Calculated area 
area = cv2.contourArea(contours[1])

print('area: ', area)
cv2.imshow('img', img)

#  Circumference calculation 
perimeter = cv2.arcLength(contours[1], True)
print('perimeter:', perimeter)

cv2.waitKey(0)
cv2.destroyAllWindows()

5 Polygon approximation and convex hull

findContours Contour information after contours May be too complex and not smooth , It can be used approxPolyDP Function to properly approximate the polygonal curve , This is it. Polygonal approximation of contour .

apporxPolyDP That is to say Polygon to approach contour , It's using Douglas-Peucker Algorithm （ In the method name DP）

DP The principle of the algorithm is relatively simple , The core is to constantly find the farthest point of the polygon and join it to form a new polygon , Until the shortest distance is less than the specified accuracy .

• approxPolyDP(curve, epsilon, closed[, approxCurve])

  • curve The contour to be approximated

  • epsilon namely DP The threshold used by the algorithm     threshold The smaller it is , The better the result.

  • closed Whether the contour is closed

  • The return value is the data of polygon approximation , utilize drawContours Just draw it .

Case study ：

import cv2
import numpy as np


img = cv2.imread('./hand.png')

gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

#  Two valued ,  Pay attention to it 2 Return values ,  Thresholds and results 
ret, binary = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)


#  Contour search ,  The new version returns two results ,  Outline and hierarchy ,  The old version returns 3 Parameters ,  Images ,  Outline and hierarchy 
result, contours, hierarchy = cv2.findContours(binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

#  Draw the outline ,  Be careful ,  Drawing the outline will change the original drawing 
cv2.drawContours(img, contours, 0, (0, 0, 255), 2)
#  Show the outline before polygon approximation 


#  Do polygon approximation ,  Returns a series of points on a polygon ,  That is, the outline after polygon approximation 
approx = cv2.approxPolyDP(contours[0], 20, True)
# print(type(approx))
# print(approx)
# print('--------------------------------------')
# print(contours[0])

#  Draw the outline of polygon approximation .
cv2.drawContours(img, [approx], 0, (0, 255, 0), 2)
cv2.imshow('img', img)

cv2.waitKey(0)
cv2.destroyAllWindows()

6. convex hull

The approximation polygon is the height approximation of the contour , But sometimes , We want to use a convex polygon to simplify it . A convex hull is like an approximation polygon , It's just the convex polygon on the outermost layer of the object . Convex hull refers to the complete inclusion of the original contour , And a polygon composed only of points on the contour . Every part of a convex hull is convex , That is, the straight line connecting any two points in the convex hull is inside the convex hull . In the convex hull , The internal angle of any three consecutive points is less than 180°.

convexHull(points[, hull[, clockwise[, returnPoints]]])

• points Contour data , from findContours obtain

• colckwise Draw clockwise

Case code ：

import cv2
import numpy as np


img = cv2.imread('./hand.png')

gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

#  Two valued ,  Pay attention to it 2 Return values ,  Thresholds and results 
ret, binary = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)


#  Contour search ,  The new version returns two results ,  Outline and hierarchy ,  The old version returns 3 Parameters ,  Images ,  Outline and hierarchy 
result, contours, hierarchy = cv2.findContours(binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

#  Draw the outline ,  Be careful ,  Drawing the outline will change the original drawing 
cv2.drawContours(img, contours, 0, (0, 0, 255), 2)


#  Do polygon approximation ,  Returns a series of points on a polygon ,  That is, the outline after polygon approximation 
approx = cv2.approxPolyDP(contours[0], 20, True)

#  Draw the outline of polygon approximation .
cv2.drawContours(img, [approx], 0, (0, 255, 0), 2)


#  Calculating convex hull 
hull = cv2.convexHull(contours[0])
cv2.drawContours(img, [hull], 0, (255, 0, 0), 2)

cv2.imshow('img', img)

cv2.waitKey(0)
cv2.destroyAllWindows()

 7.  Outside rectangle

Circumscribed rectangle is divided into minimum circumscribed rectangle and maximum circumscribed rectangle .

The red rectangle in the figure below is the smallest circumscribed rectangle , The green rectangle is the largest circumscribed rectangle .

• minAreaRect(points) Minimum circumscribed matrix

  • points That is, contour data  findContours obtain

  • Return a tuple , The content is a rotating rectangle (RotatedRect) Parameters of : The starting coordinates of the rectangle x,y, The width and height of the rectangle , Selection angle of rectangle .

• boundingRect(points) Maximum circumscribed matrix

  • points That is, contour data  findContours obtain

  Case code ：

import cv2
import numpy as np


img = cv2.imread('./hello.jpeg')

gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

ret, binary = cv2.threshold(gray, 150, 255, cv2.THRESH_BINARY)

result, contours, hierarchy = cv2.findContours(binary, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

#  The outermost outline is the whole image , contours[1] Represents the graphic outline in the image 
#  Notice that the returned content is a rotating rectangle ,  Contains the starting coordinates of the rectangle ,  Width, height and selection angle 
(x, y), (w, h), angle = cv2.minAreaRect(contours[1])

print(x, y)
print(w, h)
print(angle)
r = cv2.minAreaRect(contours[1])

#  Quickly rotatedrect Convert to contour data 
box = cv2.boxPoints(r)
print(box)
#  Contour must be an integer ,  It can't be a decimal ,  So convert it to an integer 
box = np.round(box).astype('int64')
print(box)
#  Draw the smallest circumscribed rectangle 
cv2.drawContours(img, [box], 0, (255, 0, 0), 2)

#  Returns the of a rectangle x,y and w,h
x,y, w, h = cv2.boundingRect(contours[1])
cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2)

cv2.imshow('img', img)
cv2.waitKey(0)
cv2.destroyAllWindows()