Public Lab Research note


The JerkPan

by cfastie | January 27, 2016 05:23 27 Jan 05:23 | #12624 | #12624

Above: A 360° panorama stitched from 40 photos taken by a KAP rig with no motors or electronics.

The new minimalist camera rig for aerial photography has a limitation. In its simplest form, it cannot take oblique photos with the camera in landscape (horizontal) position. Taking oblique photos in portrait mode is no problem. The one situation in which portrait mode might be preferable is when the photos will be stitched together into a panorama. The bigger vertical angle of view in portrait mode is desirable when only one row of photos can be taken. So if the rig could pan and point the camera in different directions, portrait mode would become a benefit instead of a limitation.

PortraitOblique.jpg
The new aerial rig with the camera in portrait mode and ready to take oblique photos.

Panning the rig is tricky without adding lots of complexity. User Hobbiestoomany on Shapeways posted a design which has the potential to allow crude panning with the addition of some small pieces of plastic. I am trying to adapt this idea for the new rig. No CAD files were available, so I started from scratch and designed the JerkPan in Sketchup. I made the first test of this sketchy system, and it looks promising. The video below describes the progress so far.

The field test in the video used landscape mode, and it will be more difficult to get continuous coverage with portrait mode. It will also be harder to take enough photos for a big panorama before the kite or balloon moves the rig too far. But making spherical panoramas is not the only goal. The potential to stitch a few oblique aerial photos together into a panorama could be a useful addition. In fact, the ability to capture photos in many directions instead of just one might make the JerkPan a desirable enhancement. But lots more testing is called for.


I did this Help out by offering feedback!


People who did this (0)

None yet. Be the first to post one!


5 Comments

I wonder how much motion it would get on a simple flight without jostling it at all. This is such a great idea, allowing for chaotic energy to get you a desired effect!

I wonder if you could get good information on the rig's behavior from the GPS data on a long flight with short intervals?

Great innovation on this. I'm immediately wondering how to make an analog without the printed parts, such as using a church key on a canning lid, etc. to get similar shape into flat pieces... Very cool Chris!

Is this a question? Click here to post it to the Questions page.

Reply to this comment...


IMG_1311.JPG

This is hacky as all getout, but the general idea of a rig for oblique panos from a kit box is one that @mathew mocked up recently, and this seemed an easy addition to let it slop around to get a panorama.

Be worth a test flight to see what happened. t

Reply to this comment...


Thanks, Tony. I suspect there is going to be great variability in the rotation depending on wind, kite, camera mass, suspension (Picavet vs pendulum), line yanking, etc. Not to mention the configuration of the particular JerkPan escapement used. Getting good data on how it all works might be as simple as recording video during a flight. Then I could just yell when I jerk the line to record kite handler input. I might want to try that only when no one else is around.

It might seem that 3D printing is a poor way to manufacture an escapement for mass sales, but maybe it's not a bad idea. A JerkPan is very easy to print and small enough to produce fairly quickly. It seems a little strange to ignore Public Lab's ten 3D printers in favor of jury rigged office supplies. For example, the camera rigs you describe in your recent research notes are ingenious and a real contribution to those doing DIY mapping. Your work prototyping these rigs helps anyone who wants to make their own. But I wonder whether a handful of the parts to make those rigs is a viable product for the Public Lab Store. Including a precisely designed and manufactured component that actually works and that most people cannot make themselves seems like a good way to add value to a product.

I think it might be possible to sell the new aluminum rig with 3D printed leg brackets, hinged platform, and JerkPan for not much more than you would sell a kit of dowels, rubber bands, cable ties, and thumbtacks. This could be a way to offer greater value to customers.

Manufacturing on a home 3D printer does not scale very well past a few hundred pieces a year. But at greater numbers maybe the manufacture can be outsourced. Anyway, worrying about selling too many kite photography rigs is something I don't bother to do anymore.

Chris

Reply to this comment...


Good points. the main thrust of developing extreme low cost rigs is to be able to add them easily to the existing Balloon kit and kite kits, offering everything you need in one box for almost nothing extra. Currently we give instructions for the soda bottle rig, but I feel sending all the parts and plans for at least a mapping configuration makes such a kit feel like a complete product. I don't anticipate this would actually add to the cost of the kit. mathew and I are set to discuss adding some flavor of this in the future.

I definitely don't view a thumbtack version as viable, compared with the mechanism you're working on. That was kinda just for fun. I do think the chopstick and cardboard mapping rig is a real solution, and the pendulum rig for mobius is usable as is (or could be) with a few adjustments.

I was more thinking that having little "hacker kits" of common parts is a good way to support development that can become plans for rigs anyone can build, using only found materials and no tools, buying nothing from us at all.

I also agree these are viable as 3d printed products, and once the design matures over some iterations, there will be opportunity to then find a more mass-market manufacturing solution. I think it's all really exciting, and you've been kicking out some amazing creations!

Reply to this comment...


Oh, I like that idea of including some parts in the mapping kits so the user can get started attaching their camera. That could be a nice touch. I guess the challenge will be to include parts that the user is comfortable trusting their camera to. But I guess no beginner is ever going to feel comfortable attaching their camera to a kite line.

Maybe the aluminum rig could be simplified so no suspension shaft is required. The top of the aluminum part can attach directly to the flexible tubing and wire (the top of the pendulum that the kite line wraps around). Then the only parts needed are: aluminum part, thumb screw, tubing, stiff wire, maybe a cotter pin. An plastic bottle can be added for camera protection.

I don't know how much it would cost to have the aluminum part fabricated. Maybe a lightweight, simple version would not cost much. A simple rig like that could add value to the mapping kits so you could charge a few extra bucks for them.

Reply to this comment...


Login to comment.