Research Corp. Technologies, Inc. v. Microsoft Corp.

627 F.3d 859, 97 U.S.P.Q. 2d (BNA) 1274, 2010 U.S. App. LEXIS 24984, 2010 WL 4971008
CourtCourt of Appeals for the Federal Circuit
DecidedDecember 8, 2010
Docket2010-1037
StatusPublished
Cited by92 cases

This text of 627 F.3d 859 (Research Corp. Technologies, Inc. v. Microsoft Corp.) is published on Counsel Stack Legal Research, covering Court of Appeals for the Federal Circuit primary law. Counsel Stack provides free access to over 12 million legal documents including statutes, case law, regulations, and constitutions.

Bluebook
Research Corp. Technologies, Inc. v. Microsoft Corp., 627 F.3d 859, 97 U.S.P.Q. 2d (BNA) 1274, 2010 U.S. App. LEXIS 24984, 2010 WL 4971008 (Fed. Cir. 2010).

Opinion

RADER, Chief Judge.

Research Corporation Technologies, Inc. (“RCT”) initiated this action against Microsoft Corporation (“Microsoft”), alleging infringement of six related patents: U.S. Patent Nos. 5,111,310 (“'310 patent”); 5,341,228 (“'228 patent”); 5,477,305 (“'305 patent”); 5,543,941 (“'941 patent”); 5,708,-518 (“'518 patent”); and 5,726,772 (“'772 patent”). The United States District Court for the District of Arizona held that certain claims of the '310 and '228 patents were invalid under 35 U.S.C. § 101. The district court further held that certain claims of the '772 and '305 patents were not entitled to claim the benefit of earlier filed applications that led to the '310 and '228 patents.

Because the '310 and '228 patents claim patent-eligible subject matter, this court reverses the district court on that point. This court also finds that claim 29 of the '305 patent deserves the earlier filing date and thus reverses the district court’s effective date ruling and remands. At the same time, this court affirms the district court’s decision that claims 4 and 63 of the '772 patent are not entitled to the earlier effective filing date.

I

RCT’s six patents relate to digital image halftoning. Digital images are, in fact, thousands of pixels arranged in arrays of rows and columns. Each pixel in a black- and-white image contains information about the gray level of the image at that particular position. A black-and-white image can have 256 shades of gray. A gray level 1 represents black and a gray level 256 represents white, with intervening numbers representing various shades of gray. For color images, a computer creates separate color-specific arrays of pixels, one array for each primary color. A color-specific array has pixels containing information about the shade level of that color at that particular position.

Digital images often show shades of gray and even a spectrum of colors. Nonetheless, computer displays and print *863 ers can only use a limited number of primary colors to display these digital images. Halftoning bridges this gap by simulating a continuous tone image through the use of dots. Halftoning techniques allow computers to present many shades and color tones with a limited number of pixel colors. These techniques place the dots of primary colors in a formation that gives the viewer the illusion of many more shades of gray or varying colors. Black-and-white printers use only black dots to give the illusion of shades of gray. Color printers typically use four primary colors — cyan, magenta, yellow, and black — to give the illusion of a spectrum of colors. Color displays often use three primary colors — red, green, and blue — to achieve the same effect. Digital halftoning technology thus allows computer displays and printers to render an approximation of an image by using fewer colors or shades of gray than the original image. For the most part, this opinion discusses halftoning technology with reference primarily to a black-and-white image with varying shades of gray, rather than a color image. The principles, however, are the same.

One method of generating a digital half-toned image is called “thresholding.” The thresholding technique uses a two-dimensional array called a “mask” that is populated with predetermined threshold numbers, which are typically between 1 and 256. The thresholds do not relate at all to the image to be halftoned. The thresholding technique compares the gray level at each pixel of the image against the threshold that corresponds to the pixel’s position. If the gray level exceeds the corresponding threshold, the pixel is turned on, ie., the computer places a “1” in the appropriate memory space. The resulting halftone image is a two-dimensional array of zeros and ones.

This imaging field uses various ways to measure the quality of a halftoning process. One method examines the “dot profiles” produced by the halftoning process. A dot profile is a halftone image that would be produced if the original image were a single shade of gray, (ie., all of the pixels have the same gray level). A dot profile is essentially a pattern of black dots on a white piece of paper. A dot profile for an original image with a high gray level would have more ones and thus more black dots than a dot profile for an image with a low gray level. Closely spaced dots are said to occur at a high frequency, and those far apart are said to occur at a low frequency. Because the human visual system is more sensitive to low frequencies than to high frequencies, viewers consider dot profiles with few low-frequency dots visually pleasing.

Another way to observe the quality of a halftone is to use a power spectrum associated with each dot profile obtained from the halftoning process. A power spectrum is a graph showing the relative frequency of dots in the dot profile at a particular gray level. The shape of the power spectrum characterizes the type of “noise” that the dot profiles exhibit. For example, a dot profile with a “white noise” exhibits a power spectrum where the frequencies are approximately equal across the graph. In contrast, a dot profile with a “blue noise” exhibits a power spectrum with primarily high frequency components and negligible low frequency components.

Figure 1 of the '310 patent shows an ideal blue noise power spectrum, which is unattainable in the real world.

*864 [[Image here]]

'310 patent fig.l. The horizontal axis represents the radial frequency, which is the reciprocal of the average spacing between the dots in the dot profile. A blue noise power spectrum has negligible frequency components below the principal frequency and high frequency components above the principal frequency. The principal frequency, fE, varies from one gray level g to another:

[[Image here]]

Id. col.6 11.25-38. In this equation, R is the distance between addressable dots on the display and the gray level g is normalized from zero to one. The principal frequency assumes its highest value for 50% gray level because at this level there are equal numbers of black and white dots. Each dot profile exhibits a power spectrum with a different radial frequency because as the gray level increases, so does the number of dots in the dot profile.

Drs. Kevin J. Parker and Theophano Mitsa, the named inventors of the six RCT patents, conceived of an improved blue noise mask. The inventors’ halftoning technique used a blue noise mask, which was stored in a computer’s memory, to carry out a pixel-by-pixel comparison of the mask to the digital image. Their half-toning technique compares the gray level of each pixel in a digital image to the corresponding threshold number in the blue noise mask to produce a halftone image.

The claimed blue noise mask has unique first and second order properties. When thresholded at A% of the maximum level, exactly A out of every 100 pixels will be greater than the threshold value. For example, when the blue noise mask is thresholded at 50% of the maximum level, exactly half of the pixels will be turned on.

Free access — add to your briefcase to read the full text and ask questions with AI

Related

Cite This Page — Counsel Stack

Bluebook (online)
627 F.3d 859, 97 U.S.P.Q. 2d (BNA) 1274, 2010 U.S. App. LEXIS 24984, 2010 WL 4971008, Counsel Stack Legal Research, https://law.counselstack.com/opinion/research-corp-technologies-inc-v-microsoft-corp-cafc-2010.