Camera Setup

The Raspberry Pi camera, in this case a Raspberry Pi Camera V2 was set up using the Picamera2 library to capture frames in real-time.

Initial Setup

The only library/configuration that I needed to make the camera to work on python in a fresh new Raspberry-pi OS (10/16/2024), with specs:

OS: Debian GNU/Linux 12 (bookworm) aarch64 
Host: Raspberry Pi 4 Model B Rev 1.4 
Kernel: 6.6.51+rpt-rpi-v8 

was:

$ sudo apt install python3-picamzero

Usage on python with Picamera2

First, import Picamera2 class.

from picamera2 import Picamera2

It provides functions for controlling the camera, configuring settings, and retrieving image frames for processing.

Then initialize and configure the camera:

# Initialize Picamera2 and configure the camera
picam2 = Picamera2()
picam2.configure(picam2.create_preview_configuration(main={"format": 'XRGB8888', "size": (640, 480)}))
picam2.start()

Here, the key points are:

• Resolution: The camera captures frames at 640x480 pixels.
• Format - XRGB8888:
  • This format is a 32-bit pixel format where each pixel consists of four 8-bit channels: X, R, G, and B - total of 4 bytes (32 bits).

Although the X channel is ignored, having it simplifies alignment (padding) and access to color channels, making XRGB8888 efficient for processing tasks.

To Capture Frames it is simple as:

frame = picam2.capture_array()

Here, frames are captured as NumPy arrays, each containing the pixel data in the XRGB8888 format.

Finnaly, at the end of the program, we use:

picam2.stop()

to halt the camera stream. Stopping the stream ensures that resources are released and allows the system to be used for other tasks without interference.

Refereces

• The Picamera2 Library