Overlaying data on images in Matplotlib - Step-by-Step Execution

Concept Flow - Overlaying data on images

Load Image

↓

Display Image

↓

Prepare Data Points

↓

Overlay Data on Image

↓

Show Final Plot

First, we load and display an image. Then, we prepare data points and overlay them on the image before showing the final combined plot.

Execution Sample

Matplotlib

import matplotlib.pyplot as plt
import numpy as np
img = np.ones((5,5))
plt.imshow(img, cmap='gray')
plt.scatter([1,3], [2,4], color='red')
plt.show()

This code loads a simple image and overlays red points on it.

Execution Table

Step	Action	Data/Variables	Result/Output
1	Import matplotlib and numpy	plt, np	Libraries ready
2	Create image array	img = 5x5 array of ones	White square image created
3	Display image with imshow	img	Image shown in grayscale
4	Prepare data points	x=[1,3], y=[2,4]	Points ready to plot
5	Overlay points with scatter	x, y, color='red'	Red points appear on image
6	Show plot	plt.show()	Final image with points displayed
7	End	-	Execution complete

💡 All steps executed, final plot displayed with image and overlaid data points

Variable Tracker

Variable	Start	After Step 2	After Step 4	Final
img	undefined	5x5 array of ones	5x5 array of ones	5x5 array of ones
x	undefined	undefined	[1, 3]	[1, 3]
y	undefined	undefined	[2, 4]	[2, 4]

Key Moments - 2 Insights

Why do the points appear on top of the image and not replace it?

Because plt.imshow draws the image first, then plt.scatter adds points on the same plot, layering them visually as shown in execution_table steps 3 and 5.

What happens if we call plt.show() before plt.scatter()?

The plot would display only the image without points, since plt.show() ends the current figure display, as seen in execution_table step 6.

Visual Quiz - 3 Questions

Look at the execution table, what is the shape of 'img' after step 2?

A An empty array

B A list of points

C A 5x5 array of ones

D A 3x3 array of zeros

💡 Hint

Check the 'Data/Variables' column in step 2 of the execution_table

At which step do the red points get added on the image?

A Step 5

B Step 3

C Step 4

D Step 6

💡 Hint

Look for the action 'Overlay points with scatter' in the execution_table

If we change the color in plt.scatter to 'blue', how would the output change?

A Points would be red

B Points would be blue

C Image color changes

D No points would show

💡 Hint

Refer to the 'color' parameter in step 5 of the execution_table

Concept Snapshot

Overlaying data on images with matplotlib:
1. Load image data (e.g., numpy array).
2. Display image using plt.imshow().
3. Prepare data points (x, y).
4. Use plt.scatter() to overlay points.
5. Call plt.show() to display combined plot.

Full Transcript

This visual execution shows how to overlay data points on an image using matplotlib. First, we import necessary libraries and create a simple image as a 5x5 array of ones. We display this image with plt.imshow. Next, we prepare data points with x and y coordinates. Using plt.scatter, we overlay these points on the image. Finally, plt.show displays the combined image and points. The execution table traces each step, showing variable states and actions. Key moments clarify why points overlay the image and the importance of calling plt.show at the right time. The quiz tests understanding of image shape, overlay step, and color changes.

Practice

1. What is the main purpose of using plt.imshow() in matplotlib when overlaying data on images?

easy

A. To save the plot as an image file

B. To display an image as the background for plotting data on top

C. To create a scatter plot of data points

D. To clear the current figure before plotting

Solution

1. Step 1: Understand the role of plt.imshow()

   This function is used to display images in matplotlib, which can serve as a background for other plots.

2. Step 2: Identify its use in overlaying data

   By showing an image first, you can then plot data points or lines on top to combine visual and numeric information.

3. Final Answer:

   To display an image as the background for plotting data on top -> Option B

4. Quick Check:

   plt.imshow() shows images [OK]

Hint: Remember: imshow shows images, not plots [OK]

Common Mistakes:

• Confusing imshow with scatter plot functions
• Thinking imshow saves images
• Using imshow to clear figures

2. Which of the following is the correct way to overlay a red scatter plot on an image using matplotlib?

easy

A. plt.scatter(x, y)
plt.show()
plt.imshow(image)

B. plt.scatter(x, y, color='red')
plt.imshow(image)
plt.show()

C. plt.imshow(image)
plt.scatter(x, y, color='red')
plt.show()

D. plt.imshow(image, color='red')
plt.scatter(x, y)
plt.show()

Solution

1. Step 1: Order of plotting matters

   The image must be shown first with plt.imshow() so that scatter points appear on top.

2. Step 2: Correct syntax for scatter color

   Use color='red' inside plt.scatter() to make points red.

3. Final Answer:

   plt.imshow(image) then plt.scatter(x, y, color='red') -> Option C

4. Quick Check:

   Image first, then scatter with color [OK]

Hint: Show image before scatter to overlay correctly [OK]

Common Mistakes:

• Plotting scatter before image hides points
• Passing color to imshow instead of scatter
• Calling plt.show() too early

3. What will be the output of the following code?

import matplotlib.pyplot as plt
import numpy as np

image = np.zeros((5,5))
x = [1, 3]
y = [2, 4]

plt.imshow(image, cmap='gray')
plt.scatter(x, y, color='blue')
plt.show()

medium

A. A white 5x5 image with two blue points at coordinates (1,2) and (3,4)

B. An error because x and y coordinates are swapped

C. A black 5x5 image with two red points at coordinates (2,1) and (4,3)

D. A black 5x5 image with two blue points at coordinates (1,2) and (3,4)

Solution

1. Step 1: Understand the image array

   The image is a 5x5 array of zeros, so it appears black with cmap='gray'.

2. Step 2: Plot scatter points

   Points at (x=1, y=2) and (x=3, y=4) are plotted in blue on top of the image.

3. Final Answer:

   A black 5x5 image with two blue points at coordinates (1,2) and (3,4) -> Option D

4. Quick Check:

   Zeros = black image, scatter color blue [OK]

Hint: Remember: imshow shows array as image, scatter uses x,y coords [OK]

Common Mistakes:

• Confusing x and y coordinates
• Assuming zeros array is white
• Mixing up scatter point colors

4. The following code is intended to overlay a green line on an image, but the line does not appear. What is the error?

import matplotlib.pyplot as plt
import numpy as np

image = np.ones((10,10))
plt.imshow(image)
plt.plot([1, 8], [1, 8], color='green')
plt.show()

medium

A. The image is white and the green line is not visible due to default alpha

B. The plot command should be called before imshow

C. The color argument should be 'c' instead of 'color'

D. The coordinates for the line are outside the image bounds

Solution

1. Step 1: Analyze the image color

   The image is an array of ones, which appears white by default.

2. Step 2: Check line visibility

   A green line on a white background may be hard to see if the line is thin and no linewidth is set.

3. Final Answer:

   The image is white and the green line is not visible due to default alpha -> Option A

4. Quick Check:

   White background hides thin green line [OK]

Hint: Check background and line colors for visibility [OK]

Common Mistakes:

• Plotting line before image hides image
• Using wrong color argument name
• Assuming coordinates are out of bounds

5. You want to overlay a heatmap of data values on top of a grayscale image using matplotlib. Which approach correctly combines the image and heatmap with transparency so both are visible?

hard

A. plt.imshow(image, cmap='gray')
plt.imshow(data, cmap='hot', alpha=0.5)
plt.show()

B. plt.imshow(data, cmap='hot')
plt.imshow(image, cmap='gray', alpha=0.5)
plt.show()

C. plt.imshow(image, cmap='gray', alpha=0.5)
plt.imshow(data, cmap='hot')
plt.show()

D. plt.imshow(image, cmap='hot')
plt.imshow(data, cmap='gray', alpha=0.5)
plt.show()

Solution

1. Step 1: Display base grayscale image first

   Use plt.imshow(image, cmap='gray') to show the background image.

2. Step 2: Overlay heatmap with transparency

   Plot data with cmap='hot' and alpha=0.5 to make it semi-transparent over the image.

3. Final Answer:

   Show grayscale image first, then heatmap with alpha=0.5 -> Option A

4. Quick Check:

   Base image first, overlay with alpha [OK]

Hint: Show base image first, overlay heatmap with alpha transparency [OK]

Common Mistakes:

• Plotting heatmap before image hides heatmap
• Not using alpha causes full coverage
• Swapping colormaps between image and data