Wednesday, February 09, 2011

Pelican Imaging Comes Out of Stealth Mode

Pelican Imaging announced that it has developed the first prototype array camera for mobile devices. Pelican Imaging has developed a computational camera array architecture and IP with 12 pending patent applications in array optics, sensors and image reconstruction algorithms. Pelican’s camera improves upon image and video quality while allowing for thinner smartphones. New applications are also enabled by introducing features such as 3-D depth, gesture control, and the ability for users to interact with the image before and after capturing the shot.

The company also announced its Technical Advisory Board members: Dr. Marc Levoy, Dr. Shree K. Nayar and Dr. Bedabrata Pain.

What Pelican has developed represents a paradigm shift in imaging and video that has the potential to overcome many of the inherent limitations of mobile cameras,” said Dr. Shree Nayar, professor at Columbia University.

Dr. Marc Levoy, professor at Stanford University commented, “Pelican’s technology has the potential to upset the traditional tradeoff between the sensitivity and resolution of a camera and its thickness. It also brings new capabilities to cameras, including post-capture focusing, foveal imaging and programmable frame rates. We have been investigating these aspects of computational photography in our laboratory at Stanford for a number of years, through the Stanford Multi-Camera Array, which is big, slow and expensive. Pelican’s solution is small, fast and inexpensive – which makes it a very exciting technology.

Pelican's architecture relaxes key pixel design constraints, thereby enhancing pixel performance beyond those in legacy cameras. I am equally excited about its potential for early and cost-effective adoption of new sensor technologies. The next generation of mini- and micro-cameras is upon us,” said Dr. Bedabrata Pain, CEO of Edict Inc.

From the company site it looks like Pelican's idea is to use many small lenses instead of a big one (something similar to Newport Imaging?):


Also, there is a nice article about Pelican's approach at Gigaom.

Thanks to OG for sending me the info!

17 comments:

  1. Looks interesting. Are there sensor innovations involved or just a camera configuration and software?

    I also wonder about the low light performance with all that computation.

    At the very least, sounds good for several niche markets even if not across the board.

    It is good to see new stuff!!

    ReplyDelete
  2. The rendering is not good for a rendering company. The cube does not in the camera body as shown. One is a cube with sharp corners, while the other is a cube with filleted sides.

    ReplyDelete
  3. Is this device used by PetaPixel or do they have their own device?
    http://www.petapixel.com/2010/09/23/the-first-plenoptic-camera-on-the-market/

    ReplyDelete
  4. are the 'many' lenses built directly into the chip's package (BGA), where you tradionally have the glass lid?

    ReplyDelete
  5. this should be another funding blackhole for US venture capitalists.

    ReplyDelete
  6. http://www.ptgrey.com/products/profusion25/index.asp

    ReplyDelete
  7. "are the 'many' lenses built directly into the chip's package (BGA), where you tradionally have the glass lid?"

    Yes. They will use Wafer level camera array plus their own custom interface chip, plus wafer level optics. Lots of investment needed to make cost effective.

    ReplyDelete
  8. It's one thing to have well-regarded outside experts as technology advisors, but I think using them for this kind of marketing is a bad move. If they're paid endorsers, their comments are meaningless. Of course they're going to give wonderful reviews.

    That said, the big challenge for this company, in my opinion, is going to be securing strong patents. Probably a best strategy would be to get a product out the door, build a customer base, and make a financial exit before any applications issue as patents and get subjected to serious scrutiny.

    I wasn't able to find any published applications at the US PTO web site that have the company as assignee, which suggests they were filed less than 18 months ago. Published application 20080080028 from 2006 lists Kartik Venkataraman as an inventor and Micron as assignee. I believe its first claim, if allowed, would cover the imaging module in the Pelican Camera shown in the image above if the underlying arrays happened to have a microlens structure.

    In general, the impression I get is that the principals were working on some ideas at Micron, and are working on substantially the same ideas at their start-up company. If I were a later investor, I would insist on seeing all of the company's patent applictions and having the best objective patent attorneys I could find go over them thoroughly before putting any money in.

    ReplyDelete
  9. @ CDM. I totally agree with your comments. The statements from these so-called experts are meaningless if only for the reasons you mention.

    ReplyDelete
  10. VC do allot of Patent Due Diligence before investing $17M. I would expect the camera module to have a regular MIPI I/F

    ReplyDelete
  11. consumers will not care of the technology innovations inside their phone camera. They want nice image quality at reasonable price. Can they satisfy this basic requirement with a sounding demo device?

    ReplyDelete
  12. If I VC, I take calculated risk today. Risk bar high for me.

    ReplyDelete
  13. good post CDM. But, remember that publishing patent applications is an option, not a requirement.
    Perhaps also there is an agreement in place with Micron to license some initial technology. Especially possible if Micron was not interested in pursuing it.
    The bottom line is do they offer a compelling advantage? Thinness? 3D? Computed focus? It seems possible to me.
    Heh, and I am NOT affiliated in any way with this company.

    ReplyDelete
  14. Computational Imaging technology using lenses "arrays" is based on Wafer scale optics infrastructure such as provided by Anteryon, Heptagon, Tessera and Nemotek.

    However, the image sensor design must be completly changed with small arrays of lower number of total effective pixels... at the end enabling to reconstruct a 8 to 12 MPixel image from lower number of effective photo-diodes. It will take time to anticipate these changes in IS design. This said, once this will be done, it will enable to stack vertically a simple optical array element to lead to a real solid-state camera with auto-focus and 3-D sensing capabilities. But the key driver is thickness, simplificaction of the supply chain and cost of the total optical module. Application could be smart-phones as well as new generation of DSC cameras (< 100$ total price) for emerging markets such as china, india, etc...

    ReplyDelete
  15. Is there any news from refocusing imaging? It seems like they have been pursuing this "computational imaging technology using lenses array" for a while already. Like what you have said, their approach suffers a huge loss in resolution.

    ReplyDelete
  16. Eric, the option not to public a patent application is obsolete. This option created effectively a lot of "under water" patents. Now all the patent applications should be published after a certain delay.

    ReplyDelete
  17. http://www.fenwick.com/docstore/publications/IP/IP_Articles/Patent_Applications_Published.pdf

    ReplyDelete

All comments are moderated to avoid spam.