avtcam

avtcam is an easy to use Python camera interface for NVIDIA Jetson or other embedded platform.

• Works with various Allied Vision Camera in Jetson Nano or other embedded platforms.

• Easily read images as numpy arrays with image = camera.read()

• Set the camera to running = True to attach callbacks to new frames

avtcam makes it easy to prototype AI projects in Python, especially within the Jupyter Lab programming environment installed in https://github.com/SunnyAVT/avtcam/examples.

Setup

[Method 1 -- install with pip]

python3 -m pip install avtcam --user

[Method 2 -- install with source code]

git clone https://github.com/SunnyAVT/avtcam
cd avtcam
python3 setup.py build
sudo python3 setup.py install

avtcam is tested against a system configured with the https://github.com/SunnyAVT/avtcam/examples . Different system configurations may require additional steps.

Usage

Below we show some usage examples. You can find more in the notebooks.

Create Allied Vision camera

Call AVTCamera to use a compatible Allied Vision camera. capture_width, capture_height, and capture_fps will control the capture shape and rate that images are aquired. width and height control the final output shape of the image as returned by the read function.

from avtcam.avt_camera import AVTCamera

camera = AVTCamera(width=224, height=224, capture_width=1024, capture_height=768, capture_device=0)

camera1 = AVTCamera(width=224, height=224, capture_width=640, capture_height=480, capture_device=1)

Read

Call read() to read the latest image as a numpy.ndarray of data type np.uint8 and shape (224, 224, 3). The color format is BGR8.

image = camera.read()

The read function also updates the camera's internal value attribute.

camera.read()
image = camera.value

Callback

You can also set the camera to running = True, which will spawn a thread that acquires images from the camera. These will update the camera's value attribute automatically. You can attach a callback to the value using the traitlets library. This will call the callback with the new camera value as well as the old camera value

camera.running = True

def callback(change):
    new_image = change['new']
    # do some processing...

camera.observe(callback, names='value')

Cameras

Allied Vision Cameras

These cameras work with the AVTCamera class. Try them out by following the example notebook.

| Model | Infared | https://github.com/NVIDIA-AI-IOT/jetcam |:-------|:-----:| | EMBEDDED VISION - ALLIED VISION'S NEW ALVIUM CAMERA SERIES | | MACHINE VISION CAMERAS |

Prerequisite

Installing Vimba SDK

For Windows:

• Download and launch the Vimba SDK installer:
  • Select "Custom Selection".
  • Select (at least) the following options:
    • A transport layer that matches your hardware (e.g. "Vimba USB Transport Layer" for USB cameras):
      • Core components.
      • Register GenICam Path variable.
    • Vimba SDK:
      • Core components.
      • Register environment variables.
      • C API runtime components.
      • C API development components.
      • Driver Installer.
      • Vimba Viewer.
• Run VimbaDriverInstaller.exe and install the relevant driver.
• Test the driver installation by running VimbaViewer.exe.

For other OS's see Vimba's download page.

Installing Pymba

For Python 3 install Pymba via PIP.

pip install pymba

See also

• jetson_nano_avtcamera_tracking - Based on jetson nano detection and tracking project, develop a AVT camera version tracking project.
• jetcam - JetCam is an easy to use Python camera interface for NVIDIA Jetson.