Extract images from video in Python

In this tutorial, we will learn how to extract images from a video file using Python and the OpenCV library.

Prerequisites

• Python 3.x
• OpenCV library for Python

Reading a Video File

First, we need to read the video file using the cv2.VideoCapture() function. This function returns a VideoCapture object, which we can use to read the frames of the video.

import cv2

# Open the video file
video = cv2.VideoCapture('video.mp4')

Extracting Frames from the Video

To extract the frames from the video, we can use a while loop and the VideoCapture.read() method. This method returns a tuple containing a boolean value indicating whether the frame was successfully read and the frame itself.

# Read the frames from the video
while True:
    # Read the next frame
    success, frame = video.read()

    # Check if we reached the end of the video
    if not success:
        break

    # Do something with the frame
    # ...

Saving the Frames as Images

To save the frames as images, we can use the cv2.imwrite() function. This function takes the file name and the image as arguments and saves the image to the specified file.

# Save the frame to an image file
cv2.imwrite('frame.jpg', frame)

Extracting Frames at a Specific Frame Rate

By default, the VideoCapture.read() method reads the frames of the video at the original frame rate. However, we can specify a different frame rate by using the VideoCapture.set() method to set the cv2.CAP_PROP_FPS property.

For example, to extract one frame per second from the video, we can use the following code:

# Set the frame rate to 1 frame per second
video.set(cv2.CAP_PROP_FPS, 1)

Putting It All Together

Here is the complete code for extracting frames from a video and saving them as images:

import cv2

# Open the video file
video = cv2.VideoCapture('video.mp4')

# Set the frame rate to 1 frame per second
video.set(cv2.CAP_PROP_FPS, 1)

# Set the starting frame
frame_number = 0

# Read the frames from the video
while True:
    # Read the next frame
    success, frame = video.read()

    # Check if we reached the end of the video
    if not success:
        break

    # Save the frame to an image file
    cv2.imwrite(f'frame_{frame_number}.jpg', frame)

    # Increment the frame number
    frame_number += 1

# Release the video capture object
video.release()

This code reads the frames of the video one by one, saves each frame to an image file with a name like frame_0.jpg, frame_1.jpg, and so on, and increments the frame number