RaspberryPi/Timelapse4

Also see RaspberryPi/USB Camera

Overview

For timelapses 1-3, I had multiple headaches dealing with the weak-sauce ribbon-cable Pi camera. I dumped it for a more hefty, webcam-style USB camera. Photo:

This camera is 1/2 inch in diameter.

Timelapse 4 Plan

First, let's start with what we like, and what we've accomplished:

• Several things were accomplished with RaspberryPi/Timelapse3 (example: we were able to control the LED).
• Can resurrect idea bucket, continue working on the weatherproofing ideas, now that the bottleneck has been removed (better camera)

Plan for weatherproof camera mount:

• KISS
• Mount it, use the cap that came with it, put a rubber gasket around it to protect from rain further
• Don't worry about making it bomb-proof your first time. Just get the parts working in sync.

Stuff List

Hardware list:

• raspberry pi 3
• usb jump drive (for photo storage)
• usb camera

Software list:

• script to operate camera
• necessary Python bindings for taking photos
• motion server/fwcapture?

Modifications

Modifications to the Pelican case are covered at the RaspberryPi/Weatherproof Camera Case page.

Pelican Case Modification: Success

Modifications made to the Pelican case, described at RaspberryPi/Weatherproof Camera Case, were successful. The camera lens was about 0.55 inches in diameter, and fits into the hole with very little clearance. I have more notes on the specifics of the case modification over at RaspberryPi/Weatherproof Camera Case. This is Stage 1 of the Pelican modifications. Eventually this will be full-on waterproof, but the goal for Timelapse 4 was to keep it simple, stupid.

Upgrading Image Software

I needed to upgrade the way I was grabbing images from the camera. I decided to bite the bullet and go with SimpleCV, which would provide access to more high-power image processing applications - exactly what I'm looking for.

Installing SimpleCV

To install SimpleCV on the Raspberry Pi, you will need to install scipy, and that's gonna take a while:

$ sudo apt-get install python-scipy

(Note: I don't recommend doing pip install scipy, as that does not install the proper linear algebra headers, and now you're going to spend your evening in Dependency Hell.)