Phillips Goes 3-D…

Phillips Goes 3-D…
Published October 25th, 2007 in Emerging Technology.

Get ready to dump those funky 3-D glasses…

Phillips Electronics has been working on a technology that will deliver a pseudo 3-D image using a special flat panel display - no colored specs required. Called ‘WoWvx’, this technology ties together unique display optics, a video meta-data stream that adds depth information to traditional 2-D images and some clever software processing to pull it all together.

The optics piece of WoWvx is based on a panel of lenticular lens placed on the face of the display. These lens allows the system to provide each eye of the viewer with its own distinct image. These distinct images offer a slightly different perspective of the same main image, creating the illusion of depth. This is exactly how ours eyes work in the real world - each one sees a slightly different image, which our brain combines to give us spacial perception.

blog-_d-postcard-lg.jpgThe concepts behind the optical technology used here have been around for a while. While clearly much more refined, the lens panel used here plays a similar role to the top layer used in those 3-D post cards that were popular in the 1960’s - the ones that used to show different images when you held the card at different angles. The difference - and it’s a BIG difference - is that this system can deliver these specific independent images to each eye simultaneously.

For this lens system to work, there is a hefty chunk of technology behind it that needs to figure out how to construct these two distinct images in the first place.

And that’s where the interesting stuff really kicks in…

For each frame of video, a map needs to exist defining the depth of each pixel in the frame. This is similar to something in video post production called ‘2.5-D’. In 2.5-D, flat 2-D images are extruded to create depth. These extruded elements can then be manipulated independently, and a ‘virtual camera’ can move around the extruded picture in 3-D, albeit with some limitations, to create a series of moving images that appear to have depth. This type of breakdown looks something like this:

blog-25d-images.jpg

In the Phillips system, this depth defines the distance of each pixel from the lens of the camera that shot the image. It can be calculated through various mechanisms directly from 2-D images (using methods like blur analysis for example), or captured at the same time as the image. Software is able to take these depth values and use them to derive two mathematically generated images that show slightly different perspectives on the original image. These two images are vertically interlaced on the screen, and the lenticular lens take it from there to deliver them to each eye.

And there you have it. 3-D! (sorta) …

While Phillips is currently targeting WOWvx at commercial venues, it seems like it eventually could find a place in home theaters as well. Here is a short video by Phillips advertising this system:

To see a how the meta-data around depth is represented, check out some of the videos over at WoWvx.com. With each of these videos, the left side of the screen shows the original image stream, and the right side shows it’s corresponding ‘depth map’. Lighter shades of gray indicate the corresponding pixels are closer to the camera and darker shades mean they are farther away. The two streams combine to make ‘3-D’ happen.

If this technology eventually takes off, those multi-colored ’stereo-vision’ glasses will be a thing of the past.

Hopefully the same will be true for those cheesy 3-D movies…