How I Designed a Compact Weatherproof Raspberry Pi Camera Case

Motivated by the video, Militarizing Your Backyard with Python, I was tinkering with Raspberry Pi for Squirrel recognition to protect my garden bustling with sweet peas and tomatoes. As a step towards the goal, I started first programming Raspberry Pi A+ with OpenCV and then emailing me the face of any person intruding in my backyard. However I found no suitable Pi case in the market which is compact, weatherproof and has elegant look. Fake security camera case sold on Amazon was quite cheap but broke easily. It made me to think towards designing a compact case which can be used outside to monitor the backyard garden and get alerts.


I sat down and sketched the general shape of case on back of a scratch paper. Here is how my sketch looked. I chose to stick with Square shape, though Pi A+ is a rectangular shape.


Sizing the Case

With no experience in designing plastic parts meant watching tutorials from YouTube and learning from the iterations. It took one month for me to become familiar with tools. I broke down the design work several intermediate goals, beginning with obtaining the dimensions of Pi A+, Pi Camera, WiFi adapter and micro USB power cord. Vernier with LCD display manufactured by General Tools was very handy for this work. Ubiquitously available CAD models of Pi and Pi camera also helped to verify the clearances in the design by importing them and assembling with the designed case. Many thanks to the effort by several community members of GrabCAD.


Next step was to size and fit all the parts in the given space. Mounting Pi A+ with SD card and WiFi adapter in the case was easy. Mounting the Pi camera above Pi in as small as possible was a little difficult. I chose to mount camera directly on the top case since the data cable of Pi is flexible.


Weatherproofing method of cellphone or electronic case is pretty standard and based on using a weatherproof seal around the case. It was not difficult to split the case into two parts and then add the groove for holding the seal. Similarly seal of the glass is hold together by the seal holder attached to the case. It compresses the seal and hold it in the place.

Designing of screw bosses was a totally new area for me. I could not find any reliable information source and resorted to plastic part design guidelines from several plastic manufactures and handbooks. I am providing the details of the one from the book, Injection Mold Design Engineering by Davide Kazmer. Suggested size of d was 88% of D.


I used FDM method for 3D printing in ABS plastic due to limitations of escape holes of SLS printing. This is how the camera looks after final design, 3D printing and assembly.


Weatherproofing Test

I tested camera in rain for more than 2 hours and no water entered in it. Here is the picture of it in rain.


After the test in thunderstorm, I opened the camera to see no water in case.


3D Model and BOM

You can view the 3D model and 3D print it from Pinshape. If you do not have the 3D printer then try finding a 3D printer on in your area and use it to print.

Models also available for new versions of Pi3 as described in this recent post and can be downloaded from Pinshape links, Pi 3 and ZeroW.

Camera Parts

  • Parts from
    4-40 Thread, 3/4″ Phillips flat head, total 4 (91771A113) for case
    4-40 Thread, 3/16″, total 4 (91772A105) for Pi
    M2 Size, 5MM, total 4 (92000A012) for Pi camera
    M1.6 Size, 4MM, total 4 (92000A002) for seal
    O-Ring, 13mm id, 1.5mm (5233T39) for seal
    O-Ring, 1.5mm, 1ft length (1187N11) for case and Shrink wrap (6855K14) for USB cable
  • Glass piece 17.9mm x 1.2mm, from (HLD-242646) or (SKU: 10640)
  • Glass piece 18mmx 1.5mm from, since 17.9 mm is not available.
  • Swivel mount from Videosecu with 1/4 inch threads

Here is the timelapse video recorded with this camera.

Raspberry Pi Camera: Face Recognition and Email Alert

Though Open CV installation is highly time consuming process on a Raspberry Pi A+, it has multiple uses when it comes to image recognition. Framework created by Tony in his face recognition treasure box project is modified here for a camera which sends messages whenever it finds someone is in the backyard.

How it works?

Camera records images constantly and then passes it to the Open CV for face recognition. Once it recognizes a face in the image, it sends alert to the email address with the image in the attachment. You can modify this code further to include any other object recognition such as cat etc.


You can download the code which uses OpenCV 2.4.9 from here. Before you use it, you need to generate a password for the app from Gmail.

Gmail provides the following guidelines to generate the app password for emailing directly from Raspberry Pi. Use “other” option to generate the password for the app to login to Gmail and send an email.

Timelapse with Raspberry Pi and RainBerry

Raspberry Pi is quite useful in recording timelapse videos. Personally I liked Raspberry Pi A+ version due to small form factor and low power consumption. With advent of RainBerry case it is easy to mount the Pi camera outdoor and record the time lapse video.

With just one raspivid command you can start recording the images at a set period and then stitch images to a desired video format with mencoder.

I mounted RainBerry in my backyard and recorded timelapse of 24 hrs period. Recording images at 1 minute time interval created this video. When recording video over such a long period best way is to login to Pi over vnc, start a screen session, start time lapse and detach screen. To stop login to pi, restore screen and stop timelapse.

Get your RainBerry case from Pinshape to create timelapse for either 3D print, city traffic, garden, sky or nature.


Video below is over a shorter period of 1hr approximately focusing on the sky with 1 minute image recording.