Weird patent for a bump-mapping technique.

[Panajev2001a]

PATENT ABSTRACTS OF JAPAN

(11)Publication number : 2004-094971

(43)Date of publication of application : 25.03.2004
(51)Int.CI.

G06T 15/00

G06T 3/00

(21)Application number : 2003-361630 (71)Applicant : SONY COMPUTER ENTERTAINMENT INC
(22)Date of filing : 22.10.2003 (72)Inventor : NAKAMURA ITARU
(30)Priority
Priority number : 11258047
Priority date : 10.09.1999
Priority country : JP
(54) IMAGE PROCESSOR, RECORDING MEDIUM, AND PROGRAM
(57)Abstract:
PROBLEM TO BE SOLVED: To provide bump mapping processing which brings about the feeling of materials of a mirror having irregularities on the surface and has a high effect though being simple.
SOLUTION: A gray pattern image Ib different in density by positions is synthesized with an original texture image Ia being a color image by a first synthesizing means 108 to obtain a first synthesized texture image Ie. Next, the original texture image Ia is enlarged on the basis of the direction or the like of a normal vector being an attribute OPA of a polygon Op to obtain an enlarged original texture image If. The enlarged original texture image If and a negative-positive inverted gray pattern image Id are synthesized by a second synthesizing means 112 to obtain a second synthesized texture image Ig. Finally, the second synthesized texture image Ig and the first synthesized texture image Ie are synthesized to produce a third synthesized texture image Ih. The feeling of materials of a mirror having the irregularities on the surface (metallic surface) is represented by the third synthesized texture image Ih.

http://www.ipdl.jpo.go.jp/homepg_e.ipdl

 

[Simon F]

I tried reading the machine translated "description" of the patent. It sort of sounds like (at least) part of it is bog-standard embossing but after getting that far my brain was hurting too much so I gave up :?

 

[Squeak]

Is it possible to link directly to the pattent?

PROBLEM TO BE SOLVED: To provide bump mapping processing which brings about the feeling of materials of a mirror having irregularities on the surface and has a high effect though being simple.

Sounds more like a different(?) way of doing EMBM.

 

[Guden Oden]

I'm surprised it is at all possible to patent techniques on how to use a common tool in a particular way, after all, the one doing the patenting didn't invent the tool.

It's like getting a patent on using a hammer to pound in a particular kind of nail in a particular kind of wood, with hammer, nails and wood all being commonly used in other circumstances.

Idiocy! Pure idiocy...

 

[Squeak]

I'm surprised it is at all possible to patent techniques on how to use a common tool in a particular way, after all, the one doing the patenting didn't invent the tool.

It's like getting a patent on using a hammer to pound in a particular kind of nail in a particular kind of wood, with hammer, nails and wood all being commonly used in other circumstances.

Idiocy! Pure idiocy...

I couldn't agree more.
But you're saying only common tools and techniques are used, does that mean that you have read and understood the patent? If so, I would be gratefull if you would explain how it works briefly.
EMBM without special hardware would be quite something.

 

[Simon F]

Is it possible to link directly to the pattent?

I don't think so. I think the URLs are dynamically created.
Just click on the link supplied link, choose PAJ, then number search, and use the publication number.

 

[Simon F]

I'm surprised it is at all possible to patent techniques on how to use a common tool in a particular way, after all, the one doing the patenting didn't invent the tool..

That may not be what's happening here. It's machine translated Japanese and "Patent speak" at that. It'd take some reading to really know what it's doing.

 

[Guden Oden]

EMBM without special hardware would be quite something.

Why?

Carmack talked long ago about some dude who had proved common alpha-blending was all that was needed to recreate any renderman shader for example.

What will we see next, someone patenting a particular way of writing with pens on paper? Well, Mosanto DID patent basmati rice in India, so I guess we're heading that way...

 

[Scott_Arm]

I'm surprised it is at all possible to patent techniques on how to use a common tool in a particular way, after all, the one doing the patenting didn't invent the tool.

It's like getting a patent on using a hammer to pound in a particular kind of nail in a particular kind of wood, with hammer, nails and wood all being commonly used in other circumstances.

Idiocy! Pure idiocy...

I seem to remember reading or hearing about a guy that patented the way he pushed his daughter on a swing, just to see if it could be done. That may be total BS though. Worth looking into.

 

[Simon F]

I seem to remember reading or hearing about a guy that patented the way he pushed his daughter on a swing, just to see if it could be done. That may be total BS though. Worth looking into.

No. IIRC, his son was the inventor whereas he just acted as the patent agent. I posted a link to the patent in a thread here about a month ago. It's quite amusing.

 

[Simon F]

EMBM without special hardware would be quite something.

Why?

Carmack talked long ago about some dude who had proved common alpha-blending was all that was needed to recreate any renderman shader for example.

So? What has that go to do with it? It needs dozens of passes.

Just because something can be done one way doesn't mean you can't patent a better (i.e. faster and/or cheaper etc) way of doing something. Actually, that's really the point of patents.

 

[Guden Oden]

It has to do with it's not really a secret that new effects can be accomplished with a 3D accelerator.

I guess what I'm saying is how can someone patent a technique which a 3D chip makes possible to ANYONE?

People have been texturing polys in various ways for DECADES now, there can't be anything original enough in that application worthy of patent protection there. It's something anyone could discover on purpose or by chance. Might have already for all we know.

 

[Simon F]

People have been texturing polys in various ways for DECADES now, there can't be anything original enough in that application worthy of patent protection there.

People have been writing music for thousands of years. We must have written all possible songs by now. :?

 

[davefb]

People have been texturing polys in various ways for DECADES now, there can't be anything original enough in that application worthy of patent protection there.

People have been writing music for thousands of years. We must have written all possible songs by now. :?

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

until stuff like this , you couldnt write harmonies properly . . i find that ( as a non-musician) just staggering . . it's why people used to do vocal harmonies instead of instrument ones as the humans could adjust to sound 'right' but an instrument can't dynamically retune itself ( ignoring electronic instruments i guess :-/ )
-dave-

dont suppose anyone has translated the bumpmap patent properly yet ? ?

 

[Vadi]

I think that this is different. When you write a song you don't patent it but the copyright protects you. If someone copied and pasted your shaders, your shader hierarchy complete with textures you would insist on on the copyright law.

 

[Zeross]

EMBM without special hardware would be quite something.

Why?

Carmack talked long ago about some dude who had proved common alpha-blending was all that was needed to recreate any renderman shader for example.

What will we see next, someone patenting a particular way of writing with pens on paper? Well, Mosanto DID patent basmati rice in India, so I guess we're heading that way...

Wow it wasn't exactly that : the dude in question is Mark Peercy from SGI and you're simplifying quite a bit by saying that alpha blending was enough. Here is the quote from Carmack :

Mark Peercy of SGI has shown, quite surprisingly, that all Renderman surface shaders can be decomposed into multi-pass graphics operations if two extensions are provided over basic OpenGL: the existing pixel texture extension, which allows dependent texture lookups (matrox already supports a form of this, and most vendors will over the next year), and signed, floating point colors through the graphics pipeline.

So you need flexible dependant texture reads and signed floating point colors. Something we only have since the R300 (and NV40 if you take in consideration the fact that he precised "through the graphics pipeline" which means FP FrameBuffer too). The basic requirement for EMBM is dependant texture reads so being able to do it without specific hardware would be quite an achievment IMO.