Python Programming - Reading & Resizing Videos in Google Colab Using OpenCV

Question

I used the below code to read and display a video file in Google Colab. However, I do not understand how to resize it to a specific size. Any help would be highly appreciated!

from IPython.display import HTML

from base64 import b64encode

video_path = '/Videos/dog.mp4'

mp4 = open(video_path, "rb").read()

data_url = "data:video/mp4;base64," + b64encode(mp4).decode()

HTML(f"""

<video width=400 controls><source src="{data_url}" type="video/mp4">

</video>""")

Answer 1

You asked for an answer. But without knowing how to get the answer, that's not worth a lot. It's more valuable to learn how to solve it for yourself, without having to pay someone, and without having to wait for a tutor to respond to you.

I assume you understand python basics, and how to work with Google Colab.

What you do it write a search phrase for your favorite internet search engine. Something like "resize video OpenCV". When I google that phrase, I got several dozen hits for web pages and YouTube videos, involving excellent websites such as GeeksForGeeks, Stack Overflow, and the OpenCV website. I even found the Wyzant web page for the answer above!

If you don't know the Google Colab environment, you could use your favorite internet search engine with a search phrase such as "introduction to Google Colab", "the basics of Google Colab", "Google Colab for beginners", or "Google Colab tutorial".

If you think you will be doing a fair amount of work with OpenCV, (this will be a bit repetitive) you can search for "introduction to OpenCV", "the basics of OpenCV", "OpenCV for beginners", or "OpenCV tutorial". You can also just go to the OpenCV website and search for introduction and tutorial web pages.

The websites for major Python modules such as OpenCV have large amounts of documentation, with tutorial and introduction sections, because the people who write these modules want excellent documentation, for two purposes:

1) to make their Python module more popular

2) to ensure that potential users don't ask them lots of basic questions

Basically, software without accessible documentation is useless.

Many of these Python modules post documentation on separate websites, such as on readthedocs.io.

Using a search engine can be done for any topic:

VBA

SQL

The potential for life on Europa (Jupiter's Moon)

Calculus

Conjugating verbs in Russian

Alexander the Great

How to fix a door knob

Answer 2

Below please find the steps.

1. read the video file using the cv2.VideoCapture() function.
2. get the dimensions of the video using the cv2.CAP_PROP_FRAME_WIDTH and cv2.CAP_PROP_FRAME_HEIGHT properties.
3. define the new dimensions that we want to resize the video to.
4. create a cv2.VideoWriter() object to save the resized video.
5. loop through each frame of the video using a while loop. Inside the loop, we read each frame using cap.read() and resize it using the cv2.resize() function.

import cv2

import numpy as np

# Read video file

cap = cv2.VideoCapture("path to your video file")

width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))

height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))

# Define the new dimensions

new_width = 640

new_height = 480

# Create a VideoWriter object to save the resized video

fourcc = cv2.VideoWriter_fourcc(*'mp4v')

out = cv2.VideoWriter('resized_video.mp4', fourcc, 20.0, (new_width, new_height))

while(cap.isOpened()):

ret, frame = cap.read()

if ret==True:

# Resize the frame inside the loop

resized_frame = cv2.resize(frame, (new_width, new_height), interpolation = cv2.INTER_LINEAR)

# Display the resized frame

cv2.imshow('Resized Frame',resized_frame)

# Write the resized frame to the output file

out.write(resized_frame)

# Press 'q' to quit

if cv2.waitKey(25) & 0xFF == ord('q'):

break

else:

break

# Release everything when done

cap.release()

out.release()

cv2.destroyAllWindows()