Skip to content

Latest commit

 

History

History

8_Image_Contours

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
README_IMG
README_IMG
 
 
.DS_Store
.DS_Store
 
 
Contour_Feature.py
Contour_Feature.py
 
 
Image_Contours.py
Image_Contours.py
 
 
README.md
README.md
 
 
feature.png
feature.png
 
 
shapes.png
shapes.png
 
 

README.md

Image Contours

Outline

  • Image Contours
  • Contour Features
    • Moments
    • Contour Area and Perimeter
    • Contour Approximation
    • Convex Hull
    • Bounding Rectangle
    • Minimum Enclosing Circle and Fitting an Ellipse
    • Extreme Points

Image Contours (Code)

In OpenCV, finding contours is like finding white object from black background. So object to be found should be white and background should be black. - (from openCV-python tutorial)

- File name: Image_Contours.py
- Input image: shapes.png
- Command Line: python Image_Contours.py -i shapes.png

Main Function: 
  - Find Contours: image, contours, hierarchy = cv2.findContours(src_img, mode, method)
    * return contour structure(list): [ [contour_1] [contour_2] [contour_3] ... ]
    
  - Draw Contours: contour_img = cv2.drawContours(src_img, contour, contour_index, color, thickness)
    * contour_index = -1, draw all the contours.

cv2.findContours() mode and method:

- mode:
  * CV_RETR_EXTERNAL: retrieves only the extreme outer contours.
  * CV_RETR_LIST: retrieves all of the contours without establishing any hierarchical relationships.
  * CV_RETR_CCOMP: retrieves all of the contours and organizes them into a two-level hierarchy. 
  * CV_RETR_TREE: retrieves all of the contours and reconstructs a full hierarchy of nested contours.   

- method:
  * CV_CHAIN_APPROX_NONE: stores absolutely all the contour points. 
  * CV_CHAIN_APPROX_SIMPLE: stores only several important points.
NOTE: 
  * FindContours() modifies the source image. So can use image.copy() to store the original image.
  * The different between two Contour approximation method show as blow:
    (1) cv2.CV_CHAIN_APPROX_SIMPLE
    (2) cv2.CV_CHAIN_APPROX_NONE

More - drawpoint function

cv2.findContours() will return contour pixels, and cv2.drawContours() will automatically connect all 
the pixels. In order to see the real return contour points, here use cv2.circle() to draw a dot on 
each contour pixel.

* Function: cv2.circle(src_img, center, radius, color, thickness)

Image Features (Code)

- File name: Contour_Feature.py
- Input image: feature.png
- Command Line: python Contour_Feature.py -i feature.png

a. Moments

Image moments help you to calculate some features like center of mass of the object, area of the object etc. Check out the wikipedia page on Image Moments - (from openCV-python tutorial)

Main Function: M = cv2.moments(contour)
Here extract useful data from moments(M) to calculate contour Centroid:
  * cx = int(M['m10']/M['m00'])
  * cy = int(M['m01']/M['m00'])

b. Contour Area and Perimeter

Main Function:
  (1) area = cv2.contourArea(contour)
  (2) perimeter = cv2.arcLength(contour, isClosed)
      - isClosed: whether the curve is closed or not.
NOTE: The unit for the Area and Perimeter is "pixel".

c. Contour Approximation

It approximates a contour shape to another shape with less number of vertices depending upon the precision(epsilon) we specify.

Main Function: approx = cv2.approxPolyDP(contour, epsilon, isClosed)

epsilon = 0.02 * cv2.arcLength(contour,isClosed)
  - epsilon: Parameter specifying the approximation accuracy. This is the maximum distance between 
             the original curve and its approximation.

d. Convex Hull (Implementation using Graham's Scan)

Main Function: hull = cv2.convexHull(contour)


NOTE: Provide implementation using Graham Scan Algorithm to find convex hull, see code link.

e. Bounding Rectangle

Main Function for "straight rectangle":
  (1) x,y,w,h = cv2.boundingRect(contour)
      * (x,y) is the top-left of the rectangle and (w,h) is its width and height.
      
  (2) bonding_box = cv2.rectangle(src_img, point_1, point_2, color, thickness)
      * point_1: Vertex of the rectangle.
      * point_2: Vertex of the rectangle opposite to point_1 .
Main Function for "rotated rectangle":
  - rectangle = cv2.minAreaRect(contour)
    * box = cv2.boxPoints(rectangle)
NOTE: use numpy to get the rotated box contours --> box = np.int0(box)

f. Minimum Enclosing Circle

Main Function for "Enclosing Circle":
 - (x,y),radius = cv2.minEnclosingCircle(contour)
   * circle_img = cv2.circle(src_img, center, radius,color, thickness)
Main Function for "Fitting an Ellipse":
  - ellipse = cv2.fitEllipse(contour)
   * ellipse_img = cv2.ellipse(src_img, ellipse, color, thickness)

g. Extreme Points

Main Function:
  - leftmost = tuple(cnt[cnt[:,:,0].argmin()][0])
  - rightmost = tuple(cnt[cnt[:,:,0].argmax()][0])
  - topmost = tuple(cnt[cnt[:,:,1].argmin()][0])
  - bottommost = tuple(cnt[cnt[:,:,1].argmax()][0])
NOTE: use cv2.circle() to draw a dot on 4 extreme points.

More Contour Properties and Functions

Code

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Reference & Acknowledgments