Color detection in python : Opencv
Color Detection in Python :
.png)
Hello and thanks for taking some time to read up my blog today. I recently learned and made a color detection program which works with a python library called 'OpenCV'. OpenCV stands for 'Computer Vision'. We could use this program to build a color tracking robot , the step needed to make this color detection code are the same in my last published blog (Face Tracking Device : Python & Arduino ) you can have a look at it : https://creat-ly.blogspot.com/2022/02/tracking-device-python-arduino-hello.html
Supplies :
1. Installing Python
2. Installing OpenCV and numpy
2. Installing OpenCV and numpy
Step 1: Installing Software supplies :
You can download python from : https://www.python.org/downloads/
Once you're done with the installation, open up your command prompt and type the following:
pip install opencv-python
This command will install both libraries needed : opencv and numpy
The following are the steps which will help you to build a real-time opencv application in python.
Step 3: Coding in Python:
Main code :
import cv2 as cv
import numpy as np
cap = cv.VideoCapture(0)
while True :
s,img = cap.read()
low = np.array([90, 80, 100])
dark = np.array([130, 255, 255])
hsv = cv.cvtColor(img,cv.COLOR_BGR2HSV)
mask = cv.inRange(hsv,low,dark)
save = cv.bitwise_and(img,img,mask=mask)
contour, heirarchy = cv.findContours(mask,cv.RETR_EXTERNAL,cv.CHAIN_APPROX_NONE)
if len(contour) != 0:
for i in contour :
if cv.contourArea(i) >500 :
x,y,w,h = cv.boundingRect(i)
cv.rectangle(save,(x,y),(x+w,y+h),(0,255,0),3)
print (x,y,w,h)
cv.imshow('video',save)
if cv.waitKey(1) & 0xFF == ord('m'):
break
cap.release()
cv.destroyAllWindows()
First of all we import the modules opencv and numpy by these two lines of code .
import cv2 as cv
import numpy as np
Videos shot with cameras are basically bunches of photos, so if we want to record a video we should make a while loop where we should write our code, then we should test if everything is good so we can store the "cap" variable in another varibale which we'll call "img".
- Defining the margin of the color we want to detect :
low = np.array([90, 80, 100])
dark = np.array([130, 255, 255])
these two variables ' low ' and 'dark' contain the values of the color we want to detect in HSV format .In my case I chosed to detect the blue color . it's up to you to chose whatever color you want to detect
HSV is a cylindrical color model that remaps the RGB primary colors into dimensions that are easier for humans to understand. Like the Munsell Color System, these dimensions are hue, saturation, and value.
- Hue specifies the angle of the color on the RGB color circle.
- Saturation controls the amount of color used.
- Value controls the brightness of the color.
The picture below shows how the HSV model looks like :
- Applying the mask on the picture :
hsv = cv.cvtColor(img,cv.COLOR_BGR2HSV)
To detect a color using open-cv we should convert the picture from BGR (Blue,Green,Red) to HSV(Hue,Saturation,Value) and we store it in " hsv " variable so we can apply the mask on the picture by using the next line .
mask = cv.inRange(hsv,low,dark)
Here we defining the range of the lowest and the darkest color we want to detect in the ' hsv ' picture .
save = cv.bitwise_and(img,img,mask=mask)
After all this steps, we define a variable called "save" where we store the result of the original image applied to the mask by the function "bitwise_and" by specifying the two input images ("img" and "mask") as the parameters which returns the merged image as the resulting image displayed as the output on the screen.
- Drawing the contours on the detected color :
contour, heirarchy = cv.findContours(mask,cv.RETR_EXTERNAL,cv.CHAIN_APPROX_NONE)
if len(contour) != 0:
for i in contour :
if cv.contourArea(i) >500 :
x,y,w,h = cv.boundingRect(i)
cv.rectangle(save,(x,y),(x+w,y+h),(0,255,0),3)
print (x,y,w,h)
Opencv offers the possibility to draw contours on different shapes , there are a lot of type of contours and
a lot of functions to use for each application, I'll leave the documentation of contours hierarchy provided by openCV
I really recomand you to take a look at it .
After setting the contour we should display the video in a window using the next line of code
cv.imshow('video',save)
This function has two parameters, 'video ' and that refers to the name of the window
and 'save' which is the variable where the result of the mask that is applied to the picture
if cv.waitKey(1) & 0xFF == ord('m'):
break
cap.release()
cv.destroyAllWindows()
Finally we should define the key that closes the window if we want to, in my case I chose the key 'm' then if the window is closed we should deallocates its memory .
The result of my code :
