Convert BGR to RGB in OpenCV for Computer Vision

Use OpenCV's cv2.cvtColor(image, cv2.COLOR_BGR2RGB) to convert an image from BGR to RGB color format.

Examples

InputA BGR image array with pixel [255, 0, 0] (blue)

OutputRGB image array with pixel [0, 0, 255] (red)

InputA BGR image array with pixel [0, 255, 0] (green)

OutputRGB image array with pixel [0, 255, 0] (green)

InputA BGR image array with pixel [0, 0, 255] (red)

OutputRGB image array with pixel [255, 0, 0] (blue)

How to Think About It

To convert BGR to RGB, you swap the first and third color channels of the image. OpenCV stores images in BGR order by default, but many libraries and displays expect RGB order. Using cv2.cvtColor with the COLOR_BGR2RGB flag handles this channel swap automatically.

Algorithm

1

Get the input image in BGR format.

2

Use OpenCV's color conversion function with the BGR to RGB flag.

3

Return the converted image in RGB format.

Code

python

import cv2
import numpy as np

# Create a sample BGR image (blue pixel)
image_bgr = np.array([[[255, 0, 0]]], dtype=np.uint8)

# Convert BGR to RGB
image_rgb = cv2.cvtColor(image_bgr, cv2.COLOR_BGR2RGB)

print('BGR pixel:', image_bgr[0,0])
print('RGB pixel:', image_rgb[0,0])

Output

BGR pixel: [255   0   0]
RGB pixel: [  0   0 255]

Dry Run

Let's trace the conversion of a single blue pixel from BGR to RGB.

1

Input BGR pixel

[255, 0, 0] (blue in BGR)

2

Apply cvtColor with COLOR_BGR2RGB

Channels swapped to [0, 0, 255]

3

Output RGB pixel

[0, 0, 255] (red in RGB)

Step	BGR Pixel	RGB Pixel
Input	[255, 0, 0]	N/A
Conversion	N/A	[0, 0, 255]

Why This Works

Step 1: Why BGR to RGB conversion is needed

OpenCV loads images in BGR order, but many systems expect RGB order, so colors appear wrong without conversion.

Step 2: How cvtColor works

cv2.cvtColor swaps the first and third channels automatically when using COLOR_BGR2RGB.

Step 3: Result of conversion

The image colors display correctly in RGB format after conversion.

Alternative Approaches

Manual channel swap using numpy

python

import numpy as np

image_rgb = image_bgr[:, :, ::-1]

This reverses the last dimension (channels) manually; it's faster but less readable than cvtColor.

Using OpenCV split and merge

python

b, g, r = cv2.split(image_bgr)
image_rgb = cv2.merge([r, g, b])

Splits channels and merges in RGB order; more verbose but explicit.

Complexity: O(n) time, O(n) space

Time Complexity

The conversion processes each pixel once, so time grows linearly with image size.

Space Complexity

A new image array is created for RGB, so space also grows linearly with image size.

Which Approach is Fastest?

Using cv2.cvtColor is optimized in OpenCV and usually faster than manual numpy slicing or splitting.

Approach	Time	Space	Best For
cv2.cvtColor	O(n)	O(n)	Reliable and optimized conversion
Numpy channel reverse	O(n)	O(n)	Fast and concise for simple swaps
cv2 split and merge	O(n)	O(n)	Explicit channel control, more verbose