I wrote program to identify a color from file and prints the name of given colors, from which color of file has the highest similarity with. I used HSV to get range of color and compare HSV to this range to get the most similar color. I try to iterate through the dictionary of colors to compare them to HSV of file I pass. However, it doesn't work and show any errors.

from pickletools import uint8
import cv2
from matplotlib import image
import numpy as np
#HSV Hue - color, Saturation - color density, Value - brightness of color. 
#use lower and upper value of a color to define range where picture will fit.
#opencv's HSV values are H(0-179)S(0-255)V(0-255)

img = cv2.imread('purple.jpg', cv2.COLOR_BGR2HSV)
#colors 
color_list= {
    "white":([0,0,50],[0,0,255]),
    "black":([0,0,0],[0,0,49]),
    "red":([0,0,225],[127,0,255]),
    "green":([40,40,40],[70,255,255]),
    "yellow":([23,41,223],[40,150,255]),
    "blue":([78,158,124],[138,255,255])
}

#get image 
img = cv2.imread('purple.jpg')
hsv = cv2.cvtColor(img,cv2.COLOR_BGR2HSV)

for color_name, (lower, upper) in color_list.items():
    lower = np.array(lower, dtype = np.uint8)
    upper = np.array(upper, dtype = np.uint8)

    mask = cv2.inRange(img,lower,upper)
    output = cv2.bitwise_and(img,img, mask = mask)

    if mask.any():
        print(f"{color_name} :{mask.sum}")

CodePudding user response：

You have the right idea. But instead of summing the pixels in mask, you need to find the number of white pixels for each color range

Here is a simple approach.

• For every color in the dictionary, obtain the mask
• The mask is a binary image (black and white image) with value of either 0 (black) and 255 (white). The white pixel means it contains the color within the specified range.
• For each color range, count the number of white pixels present in mask. The image has the corresponding color if its range has the most number of white pixels.

Additional Code:

final_color = None
final_count = None

for color_name, (lower, upper) in color_list.items():
  mask = cv2.inRange(hsv, tuple(lower), tuple(upper))

  # count non-zero pixels in mask
  count=np.count_nonzero(mask)
  
  if count > 0:
    final_count = count
    final_color = color_name
    

print("Nearest Color is {}".format(final_color))

CodePudding user response：

You need to call the cv2.cvtColor() method, like so:

import cv2

img = cv2.cvtColor(cv2.imread('purple.jpg'), cv2.COLOR_BGR2HSV)