This is a guide to creating a simple motion sensing security camera using a Raspberry Pi, USB webcam, Amazon S3, Slack. Anytime motion is detected images will be captured, uploaded to S3 and sent to a Slack channel.

• Prerequisites
  • Raspberry pi
  • USB Webcam
  • Golang
• Device Setup
  • Install dependencies
  • Configure motion
• S3 setup
• Slack setup
• Uploading images

RaspberryPi 3 is used for this build. They include builtin Wi-Fi, making them well suited for this usecase.

Configure the Raspberry Pi as follows:

• Raspbian Stretch Lite is the OS this guide was tested with
• Setup ssh access to the pi
• Connect the pi to your wireless network

Setting up ssh access and configuring the wireless is outside of the scope of this guide. I've found this guide to be helpful for that.

A USB webcam is used for this project. Any USB camera should do, I'm using a Logitech C920 Webcam.

You'll need a working Go environment on your local machine to build the code.

• Go v1.11
• dep

sudo apt-get update

# usb camera support
sudo apt-get install fswebcam

# Install motion
sudo apt-get install motion

Test the camera with fswebcam

fswebcam image.jpg

--- Opening /dev/video0...
Trying source module v4l2...
/dev/video0 opened.
No input was specified, using the first.
Adjusting resolution from 384x288 to 352x288.
--- Capturing frame...
Captured frame in 0.00 seconds.
--- Processing captured image...
Writing JPEG image to 'image.jpg'.

Now we know our webcam is at /dev/video0. If you look at image.jpg you'll see the picture it took.

Motion is used to monitor the camera.

Edit the following settings in the motion configuration file at /etc/motion/motion.conf

# Make sure the proper camera device is set
videodevice /dev/video0

# Tell motion to run as a background service
daemon on

Edit /etc/default/motion and set start_motion_daemon=yes

start_motion_daemon=yes

And restart the system

reboot

Once it comes back up open another ssh connection.

Confirm motion is running

systemctl status motion

At this point motion is running as a background service and will be automatically started after a reboot. The next section will guide you through uploading the images to AWS S3 and notifying a Slack channel.

1. Create an S3 bucket, noting the name and region
2. Create an IAM user, noting the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY
3. Create and attach an IAM policy to the user, allowing them to upload to the S3 bucket

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "cctv_upload",
            "Effect": "Allow",
            "Action": [
                "s3:PutObject",
                "s3:PutObjectAcl"
            ],
            "Resource": "arn:aws:s3:::YOUR-BUCKET-NAME/*"
        }
    ]
}

1. Create a Slack incoming webhook
2. Note the Webhook URL, e.g. https://hooks.slack.com/services/xxx/yyy/zzzzz

Edit the on_picture_save script with your AWS and Slack secrets:

# ./root/usr/local/bin/on_picture_save

SLACK_WEBHOOK_URL=https://hooks.slack.com/services/your/slack/webhook
S3_BUCKET_NAME=my-bucket-name
S3_BUCKET_REGION=us-west-2
AWS_ACCESS_KEY=xxx
AWS_SECRET_KEY=xxxxx

Build the cctv_upload program that ties the system together. It's called by the on_picture_save script when a new image is created. The program uploads the image to S3 and then sends a webhook to Slack with the image url.

make install
make build

Upload both programs to the pi

rsync -avz root/usr/local/bin/ pi@your_ip:/tmp

cp /tmp/cctv_upload /usr/local/bin/
cp /tmp/on_picture_save /usr/local/bin/

Finally, open the motion config again and configure it to call the on_picture_save script everytime it creates an image.

# /etc/motion/motion.conf
on_picture_save /usr/local/bin/on_picture_save %f

Restart motion

systemctl restart motion