GoProTests

Overview

The aim of this project is to learn how to test GoPro cameras with my dogs in order to learn what might be necessary in terms of custom equipment (mounts, harnesses, etc.) to use them with different species of animals used in therapy. The goal is to get my hands on the equipment and run some simple pilot tests with my dogs, rescued greyhounds Boom and D.

Progress

  • Equipment purchased: GoPro Hero4 Silver with waterproof casing, tripod mount, Fetch dog harness
  • January 12
    • brought my greyhound Boom to the Hacklab to try out GoPro equipment. Equipment worked perfectly, but Boom did not enjoy using it
    • tried three camera angles: back-mounted and forward facing, back-mounted and backward facing, and chest-mounted
    • all angles worked relatively well, and were largely unobstructed. Chest-mounted was quite low, and Boom’s shadow was visible. Back-mounted and backward facing was the most unobstructed view, but perhaps least interesting in terms of perspective. May have been more interesting if Boom had been more excitable, because tail.
    • would like to try again when Boom has more energy, also with an extension to make the back-mount slightly higher, and on D to see if the camera bothers a dog with a relative fear.
  • January 15
    • reflecting on the test from the other night, I think that my next steps will be to attempt to modify the Fetch harness to one, under-the-armpits, thick elasticized strap, rather than the three straps that currently stabilize the harness. Kind of along the same lines as a Thunder shirt strap. If necessary, with some kind of clip to attach to a collar?
    • in addition, would like to design attachments for the cameras themselves for the back strap so that they are a little higher up, and so that it is possible to mount multiple cameras – ideally to get a 360 degree view, once the camera POVs are stitched together
  • second test
    • second test was recording the commute from my house to the lab with the Fetch harness on D
    • Fetch harness did not bother D at all, in the ways that it bothered Boom
    • realized that I will need to make some kind of coding book for image analysis. For example, I would like to recognize at what points in the video footage some kinds of repetitive actions are taking place (ie, dog sniffing a surface, dog urinating, etc). This could probably be achieved by filming the dogs making the video, and coding what is happening at different time markers, then comparing with other videos for consistency.
  • March – presentation at the SCDS
    • note to check in w colleagues around quant analysis of biometric data once sensors are up and running
Print Friendly, PDF & Email