如何将此 HSV 转换为 RGB

Question

How to convert this HSV into RGB I would like it using cv2.color cv2.COLOR_BGR2RGB but its not same result.

This is the image i've used enter image description here

Heres' the result that I want but using the cv2.COLOR_BGR2RGB but the code below is using the cv2.COLOR_BGR2HSV enter image description here

import cv2
import numpy as np

## Read
img = cv2.imread("ni.jpg")

## convert to hsv
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

## mask of green (36,25,25) ~ (86, 255,255)
# mask = cv2.inRange(hsv, (36, 25, 25), (86, 255,255))
mask = cv2.inRange(hsv, (7, 25, 25), (70, 255,255))

## slice the green
imask = mask>0
green = np.zeros_like(img, np.uint8)
green[imask] = img[imask]

## save 
cv2.imwrite("green.png", green)

Answer 1

To convert an image from HSV to RGB you can do:

rgb = cv2.cvtColor(hsv, cv2.COLOR_HSV2RGB)

and to do HSV to BGR it is

bgr = cv2.cvtColor(hsv, cv2.COLOR_HSV2BGR)

You have to know that OpenCV is using BGR when reading/saving images. You can convert between RGB and BGR with cvtColor and cv2.COLOR_RGB2BGR, or cv2.COLOR_BGR2RGB.

EDIT:

However, if what you want is having a mask of green bananas (or yellow bananas), The issue is the way you defined green color : it is including a lot of other colors right now, including yellow.

What you can do with the HSV, is to only look at the first channel, the hue :

Here you can see that green and yellow can be differentiated : green bananas have pixel value roughly between 30 and 50, and yellow between 20 and 30.

You can do a mask with that. I used another library to do the cleanup of pixel we don't want. It is Scikit-image. This can be done in OpenCV as well, but it takes a bit more time...

SO here is my code :

import cv2
import numpy as np
import matplotlib.pyplot as plt
from skimage.morphology import remove_small_objects, remove_small_holes
## Read
img = cv2.imread("ni.jpg")

## convert to hsv
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
hue = hsv[:,:,0]

# plt.imshow(hue) # this show the figure in my post
# plt.show()

# mask = np.bitwise_and(hue > 20, hue < 35) # for yellow
mask = np.bitwise_and(hue > 30, hue < 50)  # for green
mask = remove_small_objects(mask, 1000)
mask = remove_small_holes(mask, 1000)

green = np.zeros_like(img, np.uint8)
green[mask] = img[mask]

## save
cv2.imwrite("green.png", green)