Panasonic GH4 array of 120 cameras makes 3D sculptures from still photos

[Andrew Reid]

Image Sensors World spotted an interesting development from Panasonic in Japan which features 120 GH4 bodies assembled in an array to capture a person from all angles before software is used to develop a photo-realistic 3D sculpture of the subject.

Read the full article

[Nrubloc]

This is a summarized translation for those interested in a few more details….

July 18^th 2015. The 3D Photo Lab Center opened at Panasonic Center in Osaka, Japan. A total of 120 GH4 camera units have been placed around the subject and shot with excellent synchronization at 1/1000sec.

The 4K pixel resolution from each of the 120 units of Lumix GH4 cameras, cover every angle to create an accurate 3D figure and combine to produce 20 million pixels of information. Color variation and information are greatly captured and it’s possible to create an instant “moment/atmosphere” by capturing color, hair, and clothing textures.

Generally, with hand scanner (capture device) it takes 10min. to get equivalent information recorded; of course it’s very difficult for people to stay still for 10 minutes. Now with this technique, it can capture in a “moment”; movement, multiple people in a group such as parent and kids, couples, etc… it’s a new way of keeping memories in form of 3D figures.

We would like you to experience how communication would/could change to this.

[maxotics]

Microsoft's photosynth technology now includes walk through, and walk around, imaging.

https://photosynth.net/preview/about

In other words, you could put a 4K camera on a wheeled dolly today, walk it around a person, and then put those images up on the site and someone could view a similar walk around.  Of course, you could just do a video.  However, with the above technology you can easily ZOOM IN and and set the speed of rotation, etc.  

Just another area where these 4K cameras will come in quite handy.

[sudopera]

If I remember correctly, Pinewood Studio in UK has similar setup with multiple Canon 100D bodies. You can see it in this BBC documentary:

http://www.bbc.co.uk/programmes/b05ys7zz

There was also some other movie magic stuff in the documentary, so I recommend watching it.

[Nrubloc]

Many studios and productions have used different multi-camera array setups for various movies, animations, gaming industry titles, etc. for quite a while now. Although that overall idea is nothing new really; what is a bit different as shown in the video, is the application of 3D printing of objects from the still images generated. An accurate and color represented physical sculpture/figure is the final product instead of a dynamic/interactive still or video as the final product. They didn't mention the software or exact 3D printer they were using, but its combined application seems the most interesting as well, rather than the usage of an array capture setup alone.

[Fatalfury]

There is a similar service in Estonia available from 2014. 

[Julian]

This is a summarized translation for those interested in a few more details….

July 18^th 2015. The 3D Photo Lab Center opened at Panasonic Center in Osaka, Japan. A total of 120 GH4 camera units have been placed around the subject and shot with excellent synchronization at 1/1000sec.

The 4K pixel resolution from each of the 120 units of Lumix GH4 cameras, cover every angle to create an accurate 3D figure and combine to produce 20 million pixels of information. Color variation and information are greatly captured and it’s possible to create an instant “moment/atmosphere” by capturing color, hair, and clothing textures.

Generally, with hand scanner (capture device) it takes 10min. to get equivalent information recorded; of course it’s very difficult for people to stay still for 10 minutes. Now with this technique, it can capture in a “moment”; movement, multiple people in a group such as parent and kids, couples, etc… it’s a new way of keeping memories in form of 3D figures.

We would like you to experience how communication would/could change to this.

​Something went wrong with your translation. It's 2 billion pixels

https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=nl&ie=UTF-8&u=http%3A%2F%2Fnews.panasonic.com%2Fjp%2Ftopics%2F2015%2F43994.html&edit-text=&act=url

4K is not relevant for this setup. They are using the photo mode and max. resolution of the sensor.

So actually it is 1,911,029,760 pixels of information (120 x 4608x3456 pixels).

Basically this could be done with any stills camera. The only challenge is to synchronise them perfectly.

[Damphousse]

​

4K is not relevant for this setup.

​Indeed.  Half a decade ago canon was using this concept with cheap Rebels... for something a bit more interesting.

 

[Damphousse]

Again from years ago...  Streets Ice Cream promotion.  I guess it was some kind of street fare or ice cream party and they let people try the Canon Rebel setup up.

 

[Andrew Reid]

Indeed.  Half a decade ago canon was using this concept with cheap Rebels... for something a bit more interesting.​

Why did they stop?

[Damphousse]

Why did they stop?

​They still make Rebels.  And even the old warmed over Rebels have more resolution and a bigger sensor than the GH4... and they can be bought for a fifth of the price.  Why anyone would use a lower resolution camera that costs five times the amount to do this is beyond me.

 

You spotted Canon's weak @$$ marketing ploy with its BS fullHD multicolored party box but you fell for Panasonic's marketing pretty hard on this one.  Anyway there are professional outfits all over the world shooting high dollar commercials with Canon "bullet time" setups.  I don't know why they would quintuple their camera budget for less resolution.

[Andrew Reid]

​They still make Rebels.

​No, I mean why did they stop being interesting.

[Nrubloc]

I agree 4K is not relevant to this setup. It was just mentioned in the video by Panasonic and as everyone knows marketing likes to still use the words "4K" as if it's some sort of futuristic magic, especially in the consumer market which the video was made for.

Of course many studios and productions over the years have used some form of different camera setup arrays, so most everyone knows it's not new to capture image/objects that way. I worked on two projects in the past that used the same technique, but frankly I was glad when movies/productions moved on from using it in obvious ways.

I however don't think the entire point of the video was to just show a camera array at work, well at least for me it wasn't the point of the video. Panasonic just used their cameras which is irrelevant to me (since they are Panasonic they would of course like to publicize themselves) to show how they can use the capture data to generate the final product... which is an accurate and color represented "physical sculpture/figure". I don't see that as a final product in the above mentioned example videos. So yeah, it's a bit of a marketing gimmick on the image capture side I suppose, but interesting to see the final object rendered/generated from the image data on the other hand.