How to Read Image Using OpenCV in Python - Simple Guide

Use cv2.imread() to read an image file in Python with OpenCV. Provide the image path as a string argument, and it returns the image as a NumPy array.

Syntax

The basic syntax to read an image using OpenCV in Python is:

• cv2.imread(filename, flags)

Where:

• filename is the path to the image file as a string.
• flags is optional and defines how the image is read (color, grayscale, etc.).

python

import cv2

image = cv2.imread('path/to/image.jpg', cv2.IMREAD_COLOR)

Example

This example reads an image file named example.jpg in color mode and prints its shape (dimensions).

python

import cv2

# Read the image in color mode
image = cv2.imread('example.jpg', cv2.IMREAD_COLOR)

# Check if image was loaded successfully
if image is not None:
    print('Image shape:', image.shape)
else:
    print('Failed to load image')

Output

Image shape: (height, width, 3)

Common Pitfalls

Common mistakes when reading images with OpenCV include:

• Using a wrong or non-existent file path, which causes cv2.imread() to return None.
• Not checking if the image was loaded before using it, leading to errors.
• Confusing color channels because OpenCV reads images in BGR order, not RGB.

python

import cv2

# Wrong way: not checking if image loaded
image = cv2.imread('wrong_path.jpg')
print(image.shape)  # This will cause an error if image is None

# Right way: check before using
image = cv2.imread('correct_path.jpg')
if image is not None:
    print(image.shape)
else:
    print('Image not found')

Output

Image not found

Quick Reference

Parameter	Description	Common Values
filename	Path to the image file	'image.jpg', 'folder/pic.png'
flags	How to read the image	cv2.IMREAD_COLOR (default), cv2.IMREAD_GRAYSCALE, cv2.IMREAD_UNCHANGED

Key Takeaways

Use cv2.imread() with the image file path to read images in OpenCV.

Always check if the image loaded successfully before processing it.

The default color mode reads images in BGR format, not RGB.

Provide the correct file path to avoid getting None as the result.

Use flags to control how the image is read (color, grayscale, etc.).