Throwable 360 degree Panoramic Ball Camera | Hacker News Reader

Throwable 360 degree Panoramic Ball Camera

jonaspfeil.de

204 pointskachhalimbu49 comments

Here's what I am waiting for: The 360-degree 3D panorama video format. Videos are recorded by holding up the torch-like camera orb which records in every direction. You no longer just watch recordings; you explore them. On a traditional TV, besides the playback controls, you control the direction to watch at any point during playback using a simple remote control. But the real experience is wearing the VR helmet which, using head-tracking, enables you to watch in every direction during the recorded movie. You can pretty much go and re-experience your vacation, even seeing things you never saw while you were there in person.

How far are we from this being possible with current tech?

http://www.nimmobay.com/media/360-video/360-video-nimmo-bay-...

This isn't 3D with a VR helmet, but is 360 degree video with ability to pan/zoom/pause at will.

http://www.cnn.com/interactive/2010/01/world/haiti.360/index...

We are nearly there. Here I am a couple months ago in Vienna with my prototype 360º video camera http://www.flickr.com/photos/ianjameswood/6103733114/in/set-... If you want more info, contact me via 360cities.net please :)

The greatest limiting factor to this becoming mainstream is the availability of cameras. Also, bandwidth and video capability is an issue: I can't download OR view 4k video on my still very capable 3 year-old macbook for example.

Now, Johnyzee I think mistakenly called it "360-degree 3d" actually, 360-degree images are 2-d, they are simply projected onto a spherical surface. They dont' however have any 3d information. Stereoscopic (i.e. "anaglyph, what people call 3-d, like what Avatar was) 360 video is another matter, and i don't think anyone has ever built that.

Paul Bourke has some notes on (static, not video) stereoscopic 360° panoramas, in his treasure-trove of stereoscopic resources:

http://paulbourke.net/miscellaneous/stereographics/stereopan...

Here's 360 video: http://www.immersivemedia.com/content/entertainment.php?clip...

I think it ran for about a hundred thousand dollars a couple of years ago.

This is what motivated me to start working on a 360 StreetView video recording of Buenos Aires. I'm hoping to start it soon. My homemade rig is far more primitive, though you can get good quality levels if you Do It Yourself ™.

You could also check the Sony Bloggie with the 360 attachment: http://www.youtube.com/watch?v=AmPO_teTK6s

You get a doughnut that you can later unwrap, but the vertical resolution, 512px, is very limited for my personal taste/need.

Edit: FOV is crucial for the immersion. Its funny how the coral reef link I provided reminds me of Quake 3 with "cg_fov" > 110

Disney shoots 360° movies (since the 1960s, I believe) using an array of 35mm cameras mounted on vehicles or suspended from aircraft and displays them in a round theater with handrails (which it recommends you hold onto because you'll fall over if you don't). I saw one of the resulting movies back in 1982 at Disneyland in Anaheim.

I've seen many of these (the lands at EPCOT in FL have quite a few). Never had an issue with feeling like I would fall over, but they are pretty cool.

Here's a 360 player: https://github.com/hi-res/360player

I love it! You could experience being a human canonball without the risks.

What about throwing it of the Empire State Building or any cliff? It'd be like base jumping without the 'chute!

Man that would be trippy.

Probably THE most exciting toy I have seen in a long long time. Can't wait to get my hands on one.

Elphel sells something similar - http://www3.elphel.com/eyesis-4pi

There is the advantage that in every photo you're going to have at least one human supplicating to the ball :)

supplicate: to ask or beg for something earnestly or humbly -- word of the day

Whereas a tripod 360 can't include the bass, won't all of these have a coy person directly below pretending to act natural like they didn't throw a ball straight up.

Great little product though.

If there are two people they can throw the ball at each other and not be directly in the center of the picture.

Or, maybe you can throw the ball in front of you even if nobody catches it: it looks pretty robust.

Nice bass != base in that comment. You definitely don't want a low-frequency rumble from your tripod affecting photos. ;)

It is on the very cusp where liberation turns to supplication.

Very cool idea. Did you notice the patent pending? What's your thought on the patentability of this "idea".

I know when I first saw just the picture of the ball camera, my first thought was, "oh, if you throw the ball in the air you could take a 360 at the apex". I then read that is what they are doing.

So, does the patent satisfy the "non-obviousness" test? It certainly seems clever, but it also seems to me highly likely that this is neither novel nor non-obvious.

I think the non-obviousness part is calculating and taking the photos at the apex of the curve because I didn't immediately get that. Also, why 36 cameras (why not 20, or 72)? I think the bar for non-obviousness in patents is ridiculously low.

Just capture steam in a cylinder, then extract the energy as the end of the cylinder expands with the steam!

"Oh, that's obvious. At least once you show it to me."

Very cool. Stitching needs a lot of improvement though.

Compared to Google Street View?

At my old lab we built something similar (camera + ball + throwable) 12 years ago: http://www.ri.cmu.edu/pub_files/1999/5/ICRA_Cyclops_99.pdf

Available electronic components back then were way larger than what's available today: we could only fit a single CCD camera in the ball. It did send a real-time video stream, though.

This + catapult = want.

Please don't mod me down for my dreams.

Would be better to lock the aperture and focus for all of the elements to the same values. Awesome idea!

Seems like a really cool gadget. The resulting pictures are really cool and remind of Google street view more than anything else. I wonder if some of the techniques Google used to stitch together street view pictures could also be used to improve the results from this camera...

Overall interesting idea but I would expect better quality from single device. Now it looks like that each camera is completely independent - at least it is very easy to see lines where photos are connected because of different amount of light captured.

This can be improved in the stitching software. The fact that you already know the position of every camera is a big advantage, so you can take shortcuts in the stitching. Right now the stitch is probably just a spherical projection with static image boundaries, based on the known camera angles. I see a bit of individual HDR effects to compensate for the small sensor's quality, but no blending or exposure correlation between the photos.

Looks like even 4π² steradian solid angle, not just 360°. Very cool stuff.

What I am looking forward to is a walkable sphere with a projection on the inside wall. You step in, walk in the direction you want to explore. Google street view would rock.

Much more usable & interesting than the Lytro, IMHO.

Indeed. Just imagine a wedding where dozens of people are tossing their balls as the bride walks down the aisle.

Shouldn't that be 41,253 deg^2, not 360 deg?

http://en.wikipedia.org/wiki/Square_degree

Yes, quite right. Although it would be preferable as 4π steradians as that is the SI unit of solid angle. http://en.wikipedia.org/wiki/Solid_angle

It'd be a lot more awesome if it were smaller. I can't imagine carrying that around at the moment. However, it is really cool.

This was a senior/masters project for some students. I am sure with more custom electronics you could fit the camera modules everything in something more baseball sized.

It seems like a small, cheap one might be a useful way to get a quick view of a 'hostile' environment; be that military, or some sort of dangerous physical environment.

Yep, balls are already used in hostile environments -- and there's a camera-ball grenade launcher, too! http://www.extremetech.com/extreme/92192-home-made-grenade-l...

Why does it have to be a throwable ball? I'd imagine a much better picture if I had the camera on an extendable stick. Perhaps offering both options. I also wish the stitching was better, Microsoft's Photosynth app does an amazing job. An option to be able to lock exposure may help the quality.

If you made it smaller (size of tennis ball?), you could give them to soldiers and instantly have a great intel gathering tool.

I'd have thought a single line of cameras along an American football would work 'better' for that, a sort of scan system¹ ...

¹ - realistically the scan would have to operate too quickly to make a proper scan.

Check out the other projects on that page also, the Augmented Reality Bridge builder using a white board is really cool: http://jonaspfeil.de/arbridgebuilder

It would seriously make me vomit and be in awe if you could play these sequences back in the AlloSphere at UCSB:

http://www.allosphere.ucsb.edu/

Mind is racing with possibilities of what this could do when combined with a water balloon slingshot. Wonder how waterproof and shockproof they have made it.

I'm sure this has the attention of the DOD. Personally I'd love to have one for while I was hiking.

I doubt it. The DoD will likely go the right of micro-UAVs.

They're not even difficult to build. Give me the budget for it and I'll have you a working prototype this weekend.

Its so cool that he lists hardware and software on the right side. (Hint: its not Haskell)

This + high-altitude weather balloon = nice panoramas.

Is that Khao Sok national park, Thailand at 1:03?

This would be great if you could tie it to a helium balloon and anchor it down with fishing line or the like. You could let it fly up 50 feet or so and capture some amazing images.

Reply on news.ycombinator.com