September, 2024 - My Awesome Moments

OpenCV findContours – What it is?

OpenCV findContours detects change in the image color and marked it as contour.

In this example using archery target face, findContours detects the outer circle, yellow circle, bullseye circle, logos on top left and bottom left.

In this case, findContours doesn’t detect the red circle contour.

The Original Image

The Image Gray

The Find Contours Image

Green color is the contours found on the image.

opencv findcontours on archery target face

The OpenCV Python Code

import os
import cv2 as cv
import numpy as np
import matplotlib.pyplot as plt


def findContours():
    root = os.getcwd()
    imgPath = os.path.join(root, 'images/target.jpg')
    img = cv.imread(imgPath)
    #change to GRAY to easily detect the contours
    imgGray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)

    ret, thresh = cv.threshold(imgGray, 127, 255, 0)
    contours, hierarchy = cv.findContours(thresh, cv.RETR_TREE, cv.CHAIN_APPROX_SIMPLE)


    imgContours = cv.drawContours(img, contours, -1, (0,255,0), 3)

    plt.figure()
    plt.subplot(121)
    plt.imshow(imgContours)
  
    plt.show()

    

if __name__ == '__main__':    
    print("Start")
    findContours()

import os

import cv2 as cv

import numpy as np

import matplotlib.pyplot as plt

def findContours():

root = os.getcwd()

imgPath = os.path.join(root, 'images/target.jpg')

img = cv.imread(imgPath)

#change to GRAY to easily detect the contours

imgGray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)

ret, thresh = cv.threshold(imgGray, 127, 255, 0)

contours, hierarchy = cv.findContours(thresh, cv.RETR_TREE, cv.CHAIN_APPROX_SIMPLE)

imgContours = cv.drawContours(img, contours, -1, (0,255,0), 3)

plt.figure()

plt.subplot(121)

plt.imshow(imgContours)

plt.show()

if __name__ == '__main__':

print("Start")

findContours()

Conclusion

findContours is an easy way to automatically detect an image shape or outline.

References:

OpenCV Python Image Perspective Transform

I wanted to learn OpenCV Perspective Transform. Tutorials that I found on the internet bit complex for me to really understand what it is all about. I have to read few tutorials before really understand it.

In layman term, perspective transform is taking a rectangle area from original image and re-project the cut rectangle to our defined size canvas.

To do this, you just need 4 corner pixel points and perspective transform will do all other pixels calculation automatically based on those 4 pixels points.

The Required Pixel Points

The corner pixels points must follow the order as shown in below image.

This is considered the source pixel points.

The destination points are where you want to re-project those source points into new plane.

The size of destination plane is:

width: 1000 pixels
height: 200 pixels

I just take the source rectangle width and height.

Image Used for the Perspective Transform

I used below image to do the perspective transform.

Code on Perspective Transform

Filename: Main.py

import os
import cv2 as cv
import numpy as np
import matplotlib.pyplot as plt

def perspectiveTransform():
    root = os.getcwd()
    imgPath = os.path.join(root, 'images/restaurant.png')
    img = cv.imread(imgPath)
    img = cv.cvtColor(img, cv.COLOR_BGR2RGB)
   
    src = np.array([[477,539], [1476,748], [428,786], [1520,905]], dtype=np.float32)

    width = 1000
    height = 200

   
    dest = np.array([[0,0], [width,0], [0,height], [width,100]], dtype=np.float32)
    
    transformation = cv.getPerspectiveTransform(src, dest)
    warpedImg = cv.warpPerspective(img, transformation, (width, height))


    plt.figure()
    plt.subplot(121)
    plt.imshow(img)
    plt.plot(src[:,0], src[:,1], 'r.')

    plt.subplot(122)
    plt.imshow(warpedImg)
    plt.plot(dest[:,0], dest[:,1], 'r.')
    plt.show()

    

if __name__ == '__main__':    
    print("Start")
    perspectiveTransform()

import os

import cv2 as cv

import numpy as np

import matplotlib.pyplot as plt

def perspectiveTransform():

root = os.getcwd()

imgPath = os.path.join(root, 'images/restaurant.png')

img = cv.imread(imgPath)

img = cv.cvtColor(img, cv.COLOR_BGR2RGB)

src = np.array([[477,539], [1476,748], [428,786], [1520,905]], dtype=np.float32)

width = 1000

height = 200

dest = np.array([[0,0], [width,0], [0,height], [width,100]], dtype=np.float32)

transformation = cv.getPerspectiveTransform(src, dest)

warpedImg = cv.warpPerspective(img, transformation, (width, height))

plt.figure()

plt.subplot(121)

plt.imshow(img)

plt.plot(src[:,0], src[:,1], 'r.')

plt.subplot(122)

plt.imshow(warpedImg)

plt.plot(dest[:,0], dest[:,1], 'r.')

plt.show()

if __name__ == '__main__':

print("Start")

perspectiveTransform()

To run the program, type at the terminal

python3 Main.py

1	python3 Main.py

Folder Structure:

Reference

OpenCV Perspective Transform Youtube

Markah Diperlukan Untuk Menang Memanah Recurve Peringkat MSSD Petaling Perdana & MSSM

Berapakah Markah Memanah Recurve Yang Diperlukan Untuk Menang Peringkat MSSD Petaling Perdana dan Peringkat MSSM untuk Lelaki Bawah 12 tahun?

Untuk daerah Petaling Perdana, U12 markah purata setiap pusingan untuk menang ialah 332 manakala peringkat MSSM ialah 352 ini berdasarkan markah MSSD Petaling Perdana 2024 dan MSSM 2024.

Markah Memanah MSSD Petaling Perdana 2024 – Lelaki Bawah 12 Tahun

Markah Memanah MSSM 2024 – Lelaki Bawah 12 Tahun