Displaying images (cv2.imshow, matplotlib) in Computer Vision - ML Experiment: Train & Evaluate

Experiment - Displaying images (cv2.imshow, matplotlib)

Problem:You want to display images correctly using two popular Python libraries: OpenCV (cv2.imshow) and Matplotlib (plt.imshow). Currently, your images either do not show up or appear with wrong colors.

Current Metrics:Images either do not display or show incorrect colors (e.g., blue and red channels swapped).

Issue:OpenCV uses BGR color format while Matplotlib uses RGB. Also, cv2.imshow requires a waitKey and destroyAllWindows call to display images properly.

Your Task

Display the same image correctly using both cv2.imshow and matplotlib.pyplot.imshow, ensuring colors appear natural and images show up without errors.

You must use the same image file for both display methods.

Do not change the image content, only adjust display code.

Use only OpenCV and Matplotlib libraries.

Hint 1

Hint 2

Hint 3

Solution

Computer Vision

import cv2
import matplotlib.pyplot as plt

# Load image using OpenCV
image_bgr = cv2.imread('sample.jpg')

# Display using OpenCV (BGR format)
cv2.imshow('OpenCV Image', image_bgr)
cv2.waitKey(0)
cv2.destroyAllWindows()

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

# Display using Matplotlib
plt.imshow(image_rgb)
plt.axis('off')
plt.title('Matplotlib Image')
plt.show()

Added cv2.waitKey(0) and cv2.destroyAllWindows() after cv2.imshow to properly display the window.

Converted image from BGR to RGB before displaying with matplotlib to fix color issues.

Used plt.axis('off') to hide axes for cleaner image display in matplotlib.

Results Interpretation

Before: Images either did not appear or showed wrong colors (blue and red swapped).

After: OpenCV window shows the image correctly in BGR format. Matplotlib shows the same image with correct colors after converting to RGB.

Different libraries use different color formats and display methods. Understanding these differences and using proper conversions and display functions ensures images show correctly.

Bonus Experiment

Try displaying a grayscale image using both cv2.imshow and matplotlib.pyplot.imshow. Ensure the image appears in grayscale without color distortions.

💡 Hint

Use cv2.imread with flag cv2.IMREAD_GRAYSCALE. For matplotlib, use cmap='gray' in plt.imshow.

Practice

1. What is the main purpose of using cv2.imshow in computer vision?

easy

A. To resize an image

B. To save an image to disk

C. To convert an image from BGR to RGB

D. To open a window that displays an image

Solution

1. Step 1: Understand the function of cv2.imshow

   cv2.imshow is used to open a new window that shows the image you provide.

2. Step 2: Differentiate from other functions

   Saving images uses cv2.imwrite, color conversion uses cv2.cvtColor, and resizing uses cv2.resize.

3. Final Answer:

   To open a window that displays an image -> Option D

4. Quick Check:

   cv2.imshow shows images in a window [OK]

Hint: cv2.imshow always opens a window to show images [OK]

Common Mistakes:

• Confusing cv2.imshow with saving or converting images
• Forgetting that cv2.imshow opens a separate window
• Thinking cv2.imshow changes image data

2. Which of the following is the correct sequence to display an image using OpenCV in Python?

easy

A. cv2.imshow(), cv2.waitKey(), cv2.destroyAllWindows()

B. cv2.waitKey(), cv2.imshow(), cv2.destroyAllWindows()

C. cv2.destroyAllWindows(), cv2.imshow(), cv2.waitKey()

D. cv2.imshow(), cv2.destroyAllWindows(), cv2.waitKey()

Solution

1. Step 1: Recall the correct order of OpenCV display functions

   First, cv2.imshow() opens the image window, then cv2.waitKey() waits for a key press, and finally cv2.destroyAllWindows() closes the window.

2. Step 2: Check the options order

   Only cv2.imshow(), cv2.waitKey(), cv2.destroyAllWindows() follows this correct sequence.

3. Final Answer:

   cv2.imshow(), cv2.waitKey(), cv2.destroyAllWindows() -> Option A

4. Quick Check:

   Display, wait, then close windows [OK]

Hint: Always call waitKey after imshow before destroying windows [OK]

Common Mistakes:

• Calling destroyAllWindows before waitKey
• Not calling waitKey causing window to close immediately
• Mixing order of functions

3. What will be the output of this code snippet?

import cv2
import matplotlib.pyplot as plt
img = cv2.imread('image.jpg')
plt.imshow(img)
plt.show()

medium

A. The image displays with correct colors

B. The image displays but colors look incorrect (blue and red swapped)

C. The code throws an error because plt.imshow cannot display images

D. The image window opens but closes immediately

Solution

1. Step 1: Understand color format difference

   OpenCV loads images in BGR format, but matplotlib expects RGB format.

2. Step 2: Effect on plt.imshow

   Displaying BGR image directly with plt.imshow causes colors to appear swapped, especially red and blue.

3. Final Answer:

   The image displays but colors look incorrect (blue and red swapped) -> Option B

4. Quick Check:

   OpenCV BGR images show wrong colors in matplotlib [OK]

Hint: Convert BGR to RGB before plt.imshow to fix colors [OK]

Common Mistakes:

• Assuming plt.imshow shows correct colors without conversion
• Confusing BGR and RGB formats
• Expecting plt.imshow to throw error on BGR images

4. You wrote this code to display an image but the window closes immediately. What is the error?

import cv2
img = cv2.imread('photo.png')
cv2.imshow('Photo', img)
cv2.destroyAllWindows()

medium

A. Missing cv2.waitKey() call after cv2.imshow()

B. cv2.destroyAllWindows() should be before cv2.imshow()

C. cv2.imread() cannot read PNG files

D. Window name 'Photo' is invalid

Solution

1. Step 1: Identify missing waitKey()

   Without cv2.waitKey(), the window opens and closes immediately because the program does not wait for a key press.

2. Step 2: Confirm other options are incorrect

   Destroying windows before showing is wrong, cv2.imread supports PNG, and window names can be any string.

3. Final Answer:

   Missing cv2.waitKey() call after cv2.imshow() -> Option A

4. Quick Check:

   Always call waitKey to pause window [OK]

Hint: Always add cv2.waitKey() after imshow to keep window open [OK]

Common Mistakes:

• Forgetting waitKey causes window to close instantly
• Thinking destroyAllWindows controls window display timing
• Assuming cv2.imread can't read PNG images

5. You want to display an image inside a Jupyter notebook using matplotlib with correct colors. Which code snippet is correct?

hard

A. import cv2
import matplotlib.pyplot as plt
img = cv2.imread('img.jpg')
plt.imshow(img)
plt.show()

B. import matplotlib.pyplot as plt
img = cv2.imread('img.jpg')
plt.imshow(img)
plt.show()

C. import cv2
import matplotlib.pyplot as plt
img = cv2.imread('img.jpg')
img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
plt.imshow(img_rgb)
plt.show()

D. import cv2
img = cv2.imread('img.jpg')
cv2.imshow('Image', img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Solution

1. Step 1: Understand color format for matplotlib display

   OpenCV loads images in BGR format, but matplotlib expects RGB, so conversion is needed.

2. Step 2: Check each option

   import cv2 import matplotlib.pyplot as plt img = cv2.imread('img.jpg') plt.imshow(img) plt.show() misses color conversion, so colors will be wrong. import cv2 import matplotlib.pyplot as plt img = cv2.imread('img.jpg') img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) plt.imshow(img_rgb) plt.show() converts BGR to RGB correctly before display. import matplotlib.pyplot as plt img = cv2.imread('img.jpg') plt.imshow(img) plt.show() misses cv2 import (NameError) and color conversion. import cv2 img = cv2.imread('img.jpg') cv2.imshow('Image', img) cv2.waitKey(0) cv2.destroyAllWindows() uses cv2.imshow which opens a separate window, not inside notebook.

3. Final Answer:

   correctly converts BGR to RGB and displays image inside notebook -> Option C

4. Quick Check:

   Convert BGR to RGB before matplotlib display [OK]

Hint: Convert BGR to RGB before plt.imshow for correct colors [OK]

Common Mistakes:

• Skipping color conversion causing wrong colors
• Using cv2.imshow inside notebooks expecting inline display
• Assuming plt.imread always works for all image types