OpenCV: Capturing and Saving Frames

Capturing Frames

Open your code-oss program, let’s begin!

If you have not connected your webcam to the Jetson Nano yet, now is a good time to do so.

Create a new Python script file in the Python-AI directory by hovering over the directory in code-oss, and click on the first icon “New File” you can name it however you want but make it something recognizable like “basicCapture.py”. The “.py” is the most important part, as this allows the code-oss Python interpreter to recognize it will be working with a valid Python script.

Now copy and paste the code below to the new Python file you just created.

import cv2
width=640
height=480

camera=cv2.VideoCapture(0)
camera.set(cv2.CAP_PROP_FRAME_WIDTH,width)
camera.set(cv2.CAP_PROP_FRAME_HEIGHT,height)

while True:
  ret,frame=camera.read()
  cv2.imshow('Camera Stream',frame)
  if cv2.waitKey(1)==ord('q'):
    break

camera.release()
cv2.destroyAllWindows()

When you’re ready, right click the code window and select Run Python File in Terminal

Breaking Down the Code

• First we import the OpenCV library so we can use its functions.

import cv2

• Then we set our screen size to 640 x 480.

width=640
height=480

You can play around with different screen resolutions like 1080×720 for instance, but in upcoming projects we will be using the smaller screen size as it will help the code to run quicker. The Larger the screen size, the more pixels are used. Computer Vision processes every single pixel, so more pixels mean more data to process.

• We create the camera object.

camera=cv2.VideoCapture(0)

I have set my camera ID to (0). Remember, your camera ID maybe different.

• Then we set our predefined width and height into the camera object.

camera.set(cv2.CAP_PROP_FRAME_WIDTH,width)
camera.set(cv2.CAP_PROP_FRAME_HEIGHT,height)

• We create a while loop to continuously loop the following code which is indented within the while loop block.
• Then we grab a frame from the camera object.
• We display the frame with OpenCV’s imshow function.
• Then we display it in a window I labelled ‘Camera Stream’.

The window can be titled anything that you wish, and we insert the captured frame into that window.

• We use OpenCV’s waitKey function within an if statement, to create a condition so we can end the program, if the ‘q’ key is pressed the code will break out of the loop.

while True:
  ret,frame=camera.read()
  cv2.imshow('Camera Stream',frame)
  if cv2.waitKey(1)==ord('q'):
    break

• We release the camera.

camera.release()

• We close the window.

cv2.destroyAllWindows()

The functions used above will be used often, please make sure you understand everything explained.