Camera Module with Raspberry Pi

This guide explains how to interface with the Raspberry Pi Camera Module using two different methods: the Command Line Interface (CLI) and Python.

Prerequisites

Ensure your camera module is correctly connected to the Raspberry Pi's CSI port and that your system is up to date.

Method 1: Using the Terminal (cli)

The modern Raspberry Pi OS uses the rpicam-apps suite (formerly libcamera).

Basic Preview

To simply open the camera and see a live preview on your monitor:

text

rpicam-still

Preview with Specific Duration

By default, the preview might close quickly or stay open depending on settings. To keep the preview open for a specific amount of time, use the -t flag followed by time in milliseconds.

text

# Open camera for 5 seconds (5000 ms)
rpicam-still -t 5000

Capturing an Image

To save an image to a specific file, use the -o (output) flag.

text

rpicam-still -o path-to-pic.jpg

Adjusting Resolution

High-resolution images take up a lot of storage space. You can resize the image using --width and --height. Check the terminal logs when running the camera to see supported resolutions.

text

# Example: Resize to 1640x1232
rpicam-still -o path-to-pic.jpg --width 1640 --height 1232

Flipping the Image

If your camera is mounted upside down or facing backwards, you can rotate the image using:

--vflip: Vertical flip
--hflip: Horizontal flip

text

rpicam-still -o path-to-pic.jpg --width 1640 --height 1232 --vflip --hflip

Recording Video

To record a video, use the rpicam-vid command. The -t flag here determines the length of the recording.

text

# Record a 10-second video (10000 ms)
rpicam-vid -t 10000 -o path-to-vid.mp4

Method 2: Using Python (`picamzero`)

For programmatic control, we will use the picamzero library. This is a beginner-friendly wrapper for the camera.

Installation

This library is not installed by default. Update your repositories and install it:

bash

sudo apt update
sudo apt install python3-picamzero

Create the Python Script

Create a new file to write your code:

text

nvim camera.py

Script: Taking a Photo

Here is a complete script to configure the camera and take a picture.

bash

from picamzero import Camera
from time import sleep

# Initialize the camera object
camera = Camera()

# Set the resolution
# Note: (1640, 1232) is a "Tuple". In Python, a Tuple allows us to store
# multiple items (width and height) in a single variable, similar to an array.
camera.still_size = (1640, 1232)

# Flip the camera if it is mounted upside down
camera.flip_camera(hflip=True, vflip=True)

# Important: Wait for the camera sensor to warm up
# (adjusts light levels and white balance)
sleep(2)

# Capture the image
# You can add a `delay` parameter inside take_photo if needed
camera.take_photo("image.jpg")

print("Photo taken successfully.")

Script: Recording Video

Below is the code to record a video segment.

dockerfile

from picamzero import Camera
from time import sleep

# Initialize the camera
camera = Camera()

# Set resolution and orientation
camera.still_size = (1640, 1232)
camera.flip_camera(hflip=True, vflip=True)

# Warm up the sensor
sleep(2)

print("Starting Video Recording...")

# Record video for a specific duration (e.g., 5 seconds)
# The first argument is the filename, the second is the duration in seconds.
camera.record_video("my_video.mp4", duration=5)

print("Video saved successfully.")

Script: Continuous Capture (data Collection)

This method is ideal for collecting datasets for Deep Learning or Machine Learning. It allows you to run code between every picture (e.g., image processing or sensor readings).

Option A: Manual Loop (Recommended for processing)

This script creates a folder and continuously takes photos until you stop it (Ctrl+C).

dockerfile

from picamzero import Camera
import os
from time import sleep

FOLDER_NAME = 'dataset_images'

# Create directory if it doesn't exist
if not os.path.exists(FOLDER_NAME):
    os.makedirs(FOLDER_NAME)

# Initialize Camera
try:
    camera = Camera()
    camera.still_size = (1640, 1232)
    camera.flip_camera(hflip=True, vflip=True)
    sleep(2) # Warmup
    print("Camera has been initialized...")
except Exception as e:
    print(f"Error initializing camera: {e}")
    exit(1)

counter = 0

try:
    while True:
        file_name = f"{FOLDER_NAME}/image_{counter}.jpg"

        # Take the photo
        camera.take_photo(file_name)
        print(f"Captured: {file_name}")

        # You can add image processing logic here before the next shot

        counter += 1
        sleep(3) # Delay between shots

except KeyboardInterrupt:
    print("\nProcess interrupted by user. Exiting...")
except Exception as e:
    print(f"Unexpected error: {e}")
finally:
    print("Cleaning up Resources")

Option B: Built-in Sequence (Simpler but Blocking)

picamzero has a built-in function called capture_sequence. It is easier to write, but it blocks execution. This means you cannot run any image processing or sensor logic between the shots; the program waits until all photos are taken before moving on.

Use this only if you just need to dump images quickly without logic in between.

ini

# Takes 10 images with a 5-second interval
# Note: You cannot run code between these shots.
camera.capture_sequence(FOLDER_NAME, num_images=10, interval=5)

This guide explains how to interface with the Raspberry Pi Camera Module using two different methods: the Command Line Interface (CLI) and Python.

Prerequisites

Ensure your camera module is correctly connected to the Raspberry Pi's CSI port and that your system is up to date.

Method 1: Using the Terminal (cli)

The modern Raspberry Pi OS uses the rpicam-apps suite (formerly libcamera).

Basic Preview

To simply open the camera and see a live preview on your monitor:

text

rpicam-still

Preview with Specific Duration

By default, the preview might close quickly or stay open depending on settings. To keep the preview open for a specific amount of time, use the -t flag followed by time in milliseconds.

text

# Open camera for 5 seconds (5000 ms)
rpicam-still -t 5000

Capturing an Image

To save an image to a specific file, use the -o (output) flag.

text

rpicam-still -o path-to-pic.jpg

Adjusting Resolution

High-resolution images take up a lot of storage space. You can resize the image using --width and --height. Check the terminal logs when running the camera to see supported resolutions.

text

# Example: Resize to 1640x1232
rpicam-still -o path-to-pic.jpg --width 1640 --height 1232

Flipping the Image

If your camera is mounted upside down or facing backwards, you can rotate the image using:

--vflip: Vertical flip
--hflip: Horizontal flip

text

rpicam-still -o path-to-pic.jpg --width 1640 --height 1232 --vflip --hflip

Recording Video

To record a video, use the rpicam-vid command. The -t flag here determines the length of the recording.

text

# Record a 10-second video (10000 ms)
rpicam-vid -t 10000 -o path-to-vid.mp4

Method 2: Using Python (`picamzero`)

For programmatic control, we will use the picamzero library. This is a beginner-friendly wrapper for the camera.

Installation

This library is not installed by default. Update your repositories and install it:

bash

sudo apt update
sudo apt install python3-picamzero

Create the Python Script

Create a new file to write your code:

text

nvim camera.py

Script: Taking a Photo

Here is a complete script to configure the camera and take a picture.

bash

from picamzero import Camera
from time import sleep

# Initialize the camera object
camera = Camera()

# Set the resolution
# Note: (1640, 1232) is a "Tuple". In Python, a Tuple allows us to store
# multiple items (width and height) in a single variable, similar to an array.
camera.still_size = (1640, 1232)

# Flip the camera if it is mounted upside down
camera.flip_camera(hflip=True, vflip=True)

# Important: Wait for the camera sensor to warm up
# (adjusts light levels and white balance)
sleep(2)

# Capture the image
# You can add a `delay` parameter inside take_photo if needed
camera.take_photo("image.jpg")

print("Photo taken successfully.")

Script: Recording Video

Below is the code to record a video segment.

dockerfile

from picamzero import Camera
from time import sleep

# Initialize the camera
camera = Camera()

# Set resolution and orientation
camera.still_size = (1640, 1232)
camera.flip_camera(hflip=True, vflip=True)

# Warm up the sensor
sleep(2)

print("Starting Video Recording...")

# Record video for a specific duration (e.g., 5 seconds)
# The first argument is the filename, the second is the duration in seconds.
camera.record_video("my_video.mp4", duration=5)

print("Video saved successfully.")

Script: Continuous Capture (data Collection)

This method is ideal for collecting datasets for Deep Learning or Machine Learning. It allows you to run code between every picture (e.g., image processing or sensor readings).

Option A: Manual Loop (Recommended for processing)

This script creates a folder and continuously takes photos until you stop it (Ctrl+C).

dockerfile

from picamzero import Camera
import os
from time import sleep

FOLDER_NAME = 'dataset_images'

# Create directory if it doesn't exist
if not os.path.exists(FOLDER_NAME):
    os.makedirs(FOLDER_NAME)

# Initialize Camera
try:
    camera = Camera()
    camera.still_size = (1640, 1232)
    camera.flip_camera(hflip=True, vflip=True)
    sleep(2) # Warmup
    print("Camera has been initialized...")
except Exception as e:
    print(f"Error initializing camera: {e}")
    exit(1)

counter = 0

try:
    while True:
        file_name = f"{FOLDER_NAME}/image_{counter}.jpg"

        # Take the photo
        camera.take_photo(file_name)
        print(f"Captured: {file_name}")

        # You can add image processing logic here before the next shot

        counter += 1
        sleep(3) # Delay between shots

except KeyboardInterrupt:
    print("\nProcess interrupted by user. Exiting...")
except Exception as e:
    print(f"Unexpected error: {e}")
finally:
    print("Cleaning up Resources")

Option B: Built-in Sequence (Simpler but Blocking)

Use this only if you just need to dump images quickly without logic in between.

ini

# Takes 10 images with a 5-second interval
# Note: You cannot run code between these shots.
camera.capture_sequence(FOLDER_NAME, num_images=10, interval=5)

Camera Module with Raspberry Pi

Prerequisites

Method 1: Using the Terminal (cli)

Basic Preview

Preview with Specific Duration

Capturing an Image

Adjusting Resolution

Flipping the Image

Recording Video

Method 2: Using Python (`picamzero`)

Installation

Create the Python Script

Script: Taking a Photo

Script: Recording Video

Script: Continuous Capture (data Collection)

Comments

Camera Module with Raspberry Pi

Prerequisites

Method 1: Using the Terminal (cli)

Basic Preview

Preview with Specific Duration

Capturing an Image

Adjusting Resolution

Flipping the Image

Recording Video

Method 2: Using Python (`picamzero`)

Installation

Create the Python Script

Script: Taking a Photo

Script: Recording Video

Script: Continuous Capture (data Collection)

Comments