Color Detection using OpenCV

For detecting a particular color using opencv we have to HSV values range for that color

For this tutorial we will be using this image to identify red color

if u know them its alright, otherwise we can find out them using opencv
let’s see how

HSV - Hue,Saturation,Value

we need to find out
min HSV-(Hue min,Saturation min,Value min) and
max HSV-(Hue max,Saturation max,Value max)
for a color we need to identify

• we will be using trackbars to find out min and max HSV values

  #creating new  window for hsv trackbars
  windowName="HSV"
  cv2.namedWindow(windowName)
  

  we need to add 6 trackbars to window

  cv2.createTrackbar(trackbarName=,windowName,value,count,onChange=onchangefunc)

• value - initial value when trackbar created
• count - max value for trackbar
• onChange - function that will execute when trackbar value is changed

  opencv has only 180 Hue values , 255 saturation and 255 values

def  onchangefunc(dummyArgument):
    #initially let onchange be empty
    pass

cv2.createTrackbar("Hue min", windowName,0,179,onchangefunc) #windowName must be the name of window we created
cv2.createTrackbar("Hue max",windowName,0,179,onchangefunc)
cv2.createTrackbar("Saturation min",windowName,0,255,onchangefunc)
cv2.createTrackbar("Saturation max",windowName,0,255,onchangefunc)
cv2.createTrackbar("Value min",windowName,0,255,onchangefunc)
cv2.createTrackbar("Value max",windowName,0,255,onchangefunc)

we successfully created trackbars

we have to create a mask based on hsv range for that color

• for that we need hsv image

  hsvImage=cv2.cvtColor(image,cv2.COLOR_BGR2HSV)
  

  

now let’s override onchangefunc

• it will be called whenever any trackbar value is changed. So, get all trackbar values in it and create a mask

• mask is an image in which our required color will be white color and remaining image in black like this

def  onchangefunc(dummyArgument):
    # get all trackbar values
    hue_min=cv2.getTrackbarPos(trackbarname="Hue min",winname=windowName)
    hue_max=cv2.getTrackbarPos(trackbarname="Hue max",winname=windowName)
    sat_min=cv2.getTrackbarPos(trackbarname="Saturation min",winname=windowName)
    sat_max=cv2.getTrackbarPos(trackbarname="Saturation max",winname=windowName)
    val_min=cv2.getTrackbarPos(trackbarname="Value min",winname=windowName)
    val_max=cv2.getTrackbarPos(trackbarname="Value max",winname=windowName)
    print((hue_min,sat_min,val_min),(hue_max,sat_max,val_max))

    # let's create mask

• adjust values of trackbars to get your required color in white and remaining in black

minHSV=numpy.array([hue_min,sat_min,val_min])
maxHSV=numpy.array([hue_max,sat_max,val_max])
mask=cv2.inRange(hsvImage,minHSV,maxHSV)  
cv2.imshow("mask",mask)

I got these values

minHSV=(141,175,0)
maxHSV=(179,255,255)

now we have to apply this mask to hsv image to get required output

#using bitwise_and operator for applying mask
finalImage=cv2.bitwise_and(image,image,mask=mask)
cv2.imshow("final image",finalImage)

Get complete source code here