In this article, we will use image processing to count the number of Objects using OpenCV in Python.

Google Colab link: https://colab.research.google.com/drive/10lVjcFhdy5LVJxtSoz18WywM92FQAOSV?usp=sharing

Module needed

• OpenCv: OpenCv is an open-source library that is useful for computer vision applications such as image processing, video processing, facial recognition, and detection, etc.
• Numpy: Numpy is a python package for scientific computing. It is a popular math library for Machine Learning. The main Object of Numpy is a multidimensional array.
• Matplotlib: Matplotlib is a Python library used for data visualization and graphical plotting of the data.

Image Used:.

Stepwise implementation

Step 1: Import required libraries. 

# Import libraries
import cv2
import numpy as np
import matplotlib.pyplot as plt

Step 2: We will read the image by using “cv2.imread(image-name)” command & then convert this image into grayscale image using “cv2.cvtColor(image-name, cv2.COLOR_BGR2GRAY)” command.

image = cv2.imread('coins.jpg')
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
plt.imshow(gray, cmap='gray')

Output:

Step 3: For counting, we have to detect the edges but before detecting the edges we have to make the image blur to avoid the noises. Use “cv2.GaussianBlur(image-name, Kernal size, std. deviation)”. 

blur = cv2.GaussianBlur(gray, (11, 11), 0)
plt.imshow(blur, cmap='gray')

Output:

Step 4:  Now we will detect edges using a canny algorithm, 2nd & 3rd parameters in cv2.canny() function are threshold values. a value between 30 & 150 are consider as an edge for this image.

canny = cv2.Canny(blur, 30, 150, 3)
plt.imshow(canny, cmap='gray')

Output:

Step 5: We can see that edges are not connected. We need to connect the edges, have to make more thiker & visible. 

dilated = cv2.dilate(canny, (1, 1), iterations=0)
plt.imshow(dilated, cmap='gray')

Output:

Step 6: Now we have to calculate the contour in the image & convert the image into RGB from BGR & then draw the contours.

(cnt, hierarchy) = cv2.findContours(
    dilated.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
cv2.drawContours(rgb, cnt, -1, (0, 255, 0), 2)

plt.imshow(rgb)

Output:

Step 7: Printing the result

print("coins in the image : ", len(cnt))

Output:

coins in the image:  5

Below is the complete implementation:

# Import libraries
import cv2
import numpy as np
import matplotlib.pyplot as plt

image = cv2.imread('coins.jpg')
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

blur = cv2.GaussianBlur(gray, (11, 11), 0)
canny = cv2.Canny(blur, 30, 150, 3)
dilated = cv2.dilate(canny, (1, 1), iterations=0)

(cnt, hierarchy) = cv2.findContours(
    dilated.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
cv2.drawContours(rgb, cnt, -1, (0, 255, 0), 2)


print("coins in the image : ", len(cnt))

Output: 

coins in the image :  5

Get the complete notebook link: click here.