IoT Cookbook > Raspberry Pi > Camera

Raspberry Pi Camera component

The Raspberry Pi Camera component allows the remote triggering of a photo via a WAMP procedure call. The photo is returned as the call result.

Raspberry Pi with camera

Trying it out

The code for this can be found in the crossbarexamples GitHub repository under iotcookbook. You need to clone this (or download it as a ZIP file).

The Pi should support most standard WebCams once you've done

sudo apt-get install fswebcam

To try it whether this works, do

fswebcam image.jpg

and then open the image.

Then open a shell for the component directory on your computer.

Start up

crossbar start

This also serves a frontend which allows you to trigger the taking of an image and displays the progress feedback as well as the image http://localhost:8080.

In camera_pi.js, add the URL of the machine on which runs:

var connection = new autobahn.Connection({
    // replace with the url of your crossbar instance
    url: "ws://<URL OF YOUR CROSSBAR INSTANCE>/ws",
    realm: "iot_cookbook"

Then get camera_pi.js onto the Raspberry Pi, e.g. by doing

scp camera_pi.js pi@<IP of your Pi>:~/

camera_pi.js is run using Node.js, so you need this installed, and additionally npm, Node.js's package managenment, and Autobahn|JS. Once you've got Node.js and npm, in the directory where camera_pi.jsis, do

npm install autobahn

(In case you get a certificate error, doing

npm config set registry

is one way to solve this, though you now no longer use https!)

We also need uuencode, which we install by doing

sudo apt-get install sharutils

Now run our camera component using Node.js

nodejs camera_pi.js

This should log

Raspberry Pi Camera starting
Procedure '' registered: 1902454329

Once this is running, in the browser frontend click on take photo. The default image of a burglar should be replaced with an image of whatever is in front of your webcam.


The component exposes a single procedure

The result is a base64 encoded image. As a default this is JPG, but can be changed (in 'camera_pi.js', see the fswebcam documentation). Due to the encoding using 'uuencode' on the Pi, header & footer need to be removed before this can be used in a data URI.

Using it

In your own project: