Free Redacted Document - District Court of Delaware - Delaware

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 1 of 46

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE

POLAROID CORPORATION, Plaintiff and Counterclaim -Defendant,
V.

C.A. No. 06-738 (SLR)

HEWLETT-PACKARD COMPANY, Defendant and Counterclaim-Plaintiff.

REDACTED

DEFENDANT HEWLETT-PACKARD COMPANY'S OPENING CLAIM CONSTRUCTION BRIEF

FISH & RICHARDSON P.C. William J. Marsden, Jr. (#2247) 919 N. Market Street, Suite 1100 Wilmington, DE 19801 Telephone: (302) 652-5070 John E. Giust Matthew E. Bernstein 12390 El Camino Real San Diego, CA 92130 Telephone: (858) 678-5070 Bradley D. Coburn 111 Congress Avenue, Suite 810 Austin, TX 78701 Telephone: (512) 472-5070 CHOATE, HALL & STEWART Daniel C. Winston Two International Place Boston, MA 02110 Telephone: (617) 248-5000 Attorneys for Defendant and Counterclaim-Plaintiff Hewlett-Packard Company January 11, 2008

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 2 of 46

TABLE OF CONTENTS Page INTRODUCTION ................................................................................................. 1 A. Technical Background ............................................................................... 1 1. 2. 3. How a Digital Image is Created from a Natural Scene ..................1 Problems with Digital Images ........................................................2 Gamma Functions and Other Transformations Used To Compensate for Problems in Image Quality ............................................................................................3 Averaging ....................................................................................... 6

4. B.

The '381 Patent and Its Prosecution History .............................................7 1. The '381 Patent ..............................................................................7 a. b. c. 2. Calculating a Local Average Luminance ........................... 8 Choosing a Gamma Function .............................................8 Applying the Gamma Transfer Function ...........................9

The Prosecution History ................................................................9

II.

THE LAW OF CLAIM CONSTRUCTION ........................................................11 A. B. C. Generally ..................................................................................................11 Means-Plus-Function Claims ...................................................................12 Preamble ..................................................................................................12

III.

HEWLETT-PACKARD'S PROPOSED CONSTRUCTIONS ...........................13 A. B. C. The Preamble And Its Terms Are Limiting .............................................13 "Continuously enhancing. ..." (claims 1,7) ............................................ 14 "Electronic Information Signals" and "Electronic Image Data Received in a Continuous Stream" (Claims 1 and 7) ......................15 1. 2. "Electronic Information Signals" Mean Luminance Signals .......................................................................16 Electronic image data received in a continuous stream of electronic information signals ......................................19 i

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 3 of 46

TABLE OF CONTENTS (cont'd)
Page

a.

Hewlett-Packard's Proposed Construction of "Continuous " Properly Relies on the Plain and Ordinary Meaning of Continuous ....................................................19 Hewlett-Packard's Proposed Construction Regarding How the "Electronic Image Data is Received" Comports with the Specification ...................................... 19

b.

D. E.

"Each signal having a value within a determinate dynamic range of values" (claims 1,7) ...................................................21 " "Averaging Claim Terms .......................................................................23 1. 2. "Averaging" ( claims 1, 7) / "Average" (claims 1, 2,7,8) ..........................................................................................23 "Means for Averaging ... and Providing an average" (claim 1 ) ........................................................................24

F. G. H. 1.

"Selecting " ( claim 7) ................................................................................28 "A select proportionate value of the dynamic range of the electronic information signals" (claim 7) ...........................................29 "Transforming " ( Claim 7) ........................................................................32 "Means for Selecting ... and for subsequently transforming" (claim 1 ) ............................................................................34 1. 2. 3. The Term is Indefinite .................................................................35 Alternatively, the Structure is Limited to Figure 4 ......................36 "Dynamic range of the electronic information signals" and "ratio of the value of the average. . " (claim 1 ) ...................................................................................37 "Low / lowest / high / highest scene light intensity levels" / "Areas of higher contrast" (claims 2, 3, 8, 9) .........................................................................38 "Determined constant " (claims 3, 9) ............................................39

4.

5. IV.

CONCLUSION ....................................................................................................40

ii

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 4 of 46

TABLE OF AUTHORITIES Page(s) FEDERAL CASES ABB Automation Inc. v. Schlumberger Resource Management Services, Inc., 254 F. Supp. 2d 475 (D. Del. 2003 ) ...................................................................................42 Amgen Inc. v. Hoechst Marion Roussel, Inc., 74,76 314 F. 3d 1313 (Fed. Cir. 2003 ) ....................................................................................74, Automotive Techs. International, Inc. v. BMW of N. America, Inc., 501 F.3d 1274 (Fed. Cir. 2007) ..........................................................................................34 B. Braun Medical, Inc. v. Abbott Laboratories, 124 F.3d 1419 (Fed. Cir. 1997) ..........................................................................................24
Biacore v. Thermo Bioanalysis Corp.,

29,30 79 F. Supp. 2d 422 (D. Del. 1999) ...............................................................................29, Bicon, Inc. v. Straumann Co., 441 F.3d 945 (Fed. Cir. 2006) ............................................................................................42 Biomedino, LLC v. Waters Techs. Corp., 490 F.3d 946 (Fed. Cir. 2007) ............................................................................................23 Cardiac Pacemakers, Inc. v. St. Jude Medical, Inc., 296 F.3d 1106 (Fed. Cir 2002) ...........................................................................................49 Cole v. Kimberly-Clark Corp., 102 F.3d 524 (Fed. Cir. 1996) ............................................................................................23 Electro Scientific Industries, Inc. v. Dynamic Details, Inc, 24,26 307 F.3d 1343 (Fed. Cir. 2002) ....................................................................................24, Faroudja Laboratories , Inc. v. Dwin Electronics, Inc., 76 F. Supp . 2d 999 (N.D. Cal. 1999 ) .................................................................................71 LizardTech, Inc. v. Earth Resource Mapping, Inc., 35,54 424 F.3d 1336 (Fed. Cir. 2005) ....................................................................................35, Medical Instrumentation & Diagnostics Corp. v. Elekta AB, 344 F.3d 1205 (Fed. Cir. 2003) ..........................................................................................71 NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282 (Fed. Cir. 2005) ..........................................................................................64 Netword, LLC v. Centraal Corp., 242 F.3d 1347 (Fed. Cir. 2001) ..........................................................................................37 Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) ..................................................................20, 35, 47, 50, 56
Poly-America, L.P. v. GSE Lining Tech., Inc.,

26,28 383 F.3d 1303 (Fed. Cir. 2004) ....................................................................................26,
iii

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 5 of 46

TABLE OF AUTHORITIES Page(s)
Porter v. Farmers Supply Serv., Inc.,

790 F.2d 882 (Fed. Cir. 1986) ......................................................................................24, 24,29 FEDERAL STATUTES 35 U.S.C. § 112 9[2 ........................................................................................................................42 35 U.S.C. § 1121 6 ............................................................................................................23, 48, 68

iv

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 6 of 46

I.

INTRODUCTION Polaroid has been, historically, a camera company. Not surprisingly, United States Patent

No. 4,829,381 ("the '381 patent") is a camera patent. In this case, as exemplified in Polaroid's claim construction positions, Polaroid seeks to broaden its patent to cover Hewlett-Packard's non-camera products, such as printers and software. Hewlett-Packard seeks to have the '381 patent interpreted as it was meant to be, based on the intrinsic record. Plaintiff Polaroid's constructions, by contrast, seek to impermissibly broaden the patent and are contrary to, and inconsistent with, the claim language, specification, and all other intrinsic evidence. So broad are Polaroid's proposed constructions that, if adopted, they would likely either render the claims of the '3 8 1 patent meaningless or broaden the patent to cover embodiments that would render the patent invalid. For these reasons, as set forth in more detail below, Defendant respectfully requests that the Court adopt Defendant's proposed constructions. A. Technical Background

The '381 patent at its most basic level is about pictures, or more particularly, making those pictures look better. The '381 patent concerns the enhancement of digital images and the use of algorithms to manipulate a digital image to emphasize certain features in the image. This enhancement is necessary because the capture and conversion of a natural scene may create problems that render an image unappealing or its details unclear. These problems are a result of the converting a natural, analog scene into a digital image. 1. How a Digital Image is Created from a Natural Scene

An imaging device, such as a digital camera, captures a scene using an array of sensors. (Joint Appendix, ("J.A."), at 8 ('381, col. 1, 11. 15-20), filed herewith.) Each sensor captures light from a small part of the scene, and then creates a digital value based on the light detected.

1

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 7 of 46

These digital values each represent the average brightness in their small part and are called "picture elements" or, more commonly, "pixels." (Rosenfeld and Kak, Digital Picture Processing, ("Rosenfeld"), Ex. A at 3.)1 In an image made up of shades of grey, a pixel can be represented with a single number - the brightness or "luminance" of the picture element. In color images, pixels are represented in different ways, two of which are relevant for the present case. First, a pixel may represent brightness using "RGB" data, where the pixel has three brightness values, relating to the red (R), green (G) and blue (B) parts of an image. (J.A. at 9 ('381, col. 3,11. 25-35).) Second, a pixel may represent brightness using what is referred to in the art as "Luminance / Chrominance." Like the RGB system, a Luminance-based representation consists of three sub-parts: Luminance (Y) and two chrominance values. (J.A. at 9 ('381, col. 3,11. 35-50).) In such a system, the luminance portion of the image is a monochrome representation of the scene's brightness, i.e. one without any color components. The collection of these pixel values, whether represented by RGB data or YCC data, result in a digital image of the natural scene. This raw data, however, often requires further processing to reduce noise or otherwise improve the quality of the image. 2. Problems with Digital Images.

One problem with digital images is that pixels have a discrete number of brightness levels that may not accurately correspond to the analog scene that was captured. For example, an putdoor scene may contain images with bright sunlight and dark shadow where the brightest value of the scene are substantially brighter than the darkest value of the scene. (J.A. at 8 ('381. col. 1, 11. 33-35).) This difference between the brightest and darkest portions is called the "dynamic range" of the scene. In a real world scene, the dynamic range may be very large, i.e., the brightest part of the scene may be many thousands of times brighter than the darkest part.

1 All exhibit citations refer to the accompanying Declaration of William J . Marsden, Jr., unless otherwise noted. 2

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 8 of 46

Alternatively, the real world scene may have a small dynamic range: a darkened theater, for example, where the light areas may only be 10 times brighter than the darkest.2 (See J.A. at 8 ('381, col. 1, 11.43-52).) This variable dynamic range of the natural environment poses a certain challenges when the real world scene is captured by an imaging device, such as a digital camera or scanner. (See, e.g., J.A. at 8 ('381, col. 1, 11. 25-45).) In an 8 bit imaging device, the pixels can range from binary values of "00000000" to "11111111," representing integer values from 0 to 255. These integer values represent the 256 levels of brightness that can be recorded by the digital imaging device. In such a case, the digital dynamic range of the imaging device is 256, ranging from a value of 255 for the brightest portions of the image and a value of zero for the darkest. When a digital imaging device with a fixed dynamic range (of, for example, 8 bits) attempts to represent part of a real world scene, it must do so using only values that range from 0 to 255. If the dynamic range of the real world scene varies by a factor of 5000, (e.g. a nighttime scene with a bright traffic light), the imaging device must attempt to represent the 5000-fold variations of the image within the 256 values that are available to it. Likewise, if a scene contains a very narrow range of luminance, it may need to be expanded out to 256 separate levels of brightness. 3. Gamma Functions and Other Transformations Used To Compensate for Problems in Image Quality.

In a perfect sensor or display, the output luminance is directly proportional to the input voltage - a voltage of 0.80 results in a luminance of 0.80. (For mathematical convenience, the range of possible voltages and luminances are typically represented with the numbers between 0 and 1 in an analog environment.) This is called a "linear" system and a graph is included below. However, televisions for example are not linear systems. The amount of light displayed by a television is not proportional to the input signal. Instead, the relationship between input

2 The noontime Sahara also has a small dynamic range: everything is very bright. 3

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 9 of 46

voltage (V) and output luminance (L) is an exponential function . Typically, L = V2'2. The exponent 2.2 in this equation is referred to as "gamma", and the effect is a gamma distortion. As shown below , an input voltage of 0.80 produces a luminance of 0.61, and an input voltage of 0.40 produces a luminance of 0.13. Linear System Gamma Distortion (e.g. normal television)
1
.80 .61

Output Luminance

Output Luminance
.13

0

.80

1

0

.40

.80

1

Input Voltage

Input Voltage

In television, gamma correction is used to "correct" gamma distortion. If a linear system is desired, the inverse of the distorting gamma can be applied to the input before sending the signal to the distorted system. For example, if a television has a gamma of 2.2, an engineer could create a linear system by first applying a gamma of 0.45 (which is 1 / 2.2). As shown below, the inverse gamma is the opposite of the gamma distortion. When applied, the distortion inherent in the display is canceled out and a linear relationship is achieved. See below:
Gamma Correction of 0.45
1, 90

Gamma Distortion of 2.2
1

bd

a,

.13 AO .90 1 0 .40 .00 1

(Computer Systems Gamma

A Linear Relationship

Graphics Development, Correction

4

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 10 of 46

Explained, Ex. B at 1-2; Deposition Transcript of Dr. Donald S. Levinstone, ("Levinstone"), Ex. C at 95). A method of adjusting luminance levels like that described above is a mathematical function that is known in the art as a "transformation". These functions can either be applied to every pixel as in the gamma correction example above, or different functions can be applied to different areas of the picture. (Rosenfeld, Ex. A at 169-73). Gamma functions are known in the art for being particularly useful in contrast adjustment. A gamma function with a low exponent stretches the portion of the dynamic range that was previously allocated to the darkest regions. For example, with a gamma exponent of 0.5, the pixels with a luminance between 0 and 0.33 in the input image, will have a luminance between 0 and 0.57 in the output image.

1

r .57

Expanded darker output values

The contrast of these darker regions is enhanced improving visibility in shadows, while the contrast of the corresponding lighter regions is compressed. This property of gamma was well known in the prior art, including in prior art cited in the '381 patent's prosecution history:3 [1]f the over all gamma characteristic is logarithmic [i.e. the exponent is less than 1], the dark picture portions will have expanded contrast, and fine dark 3 See also Levinstone, Ex. C. at 96; US Pat 4,394,688, Ex. D at 5, (col. 5,11. 13-19) (cited in Cosh, J.A. at 978). 5

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 11 of 46

or shadow detail is reproduced. Conversely, if the gamma characteristic is exponential [i.e. the exponent is greater than 1], the bright portions will have expanded contrast, and detail in brightly lit areas will be clear. (J.A. at 93 (Okada, col. 1, 1. 65 - col. 2,1. 2); id. at 88 (Okada, Fig. 2).) 4. Averaging

Processing digital images commonly involves computing a local average, often for image sharpening or noise reduction. Two methods are disclosed in the '381 patent and are relevant here: the block average and the low-pass filter. The two algorithms, both well known in the art, (J.A. at 9, ('381, col. 3, 1. 61 - col. 4, 1. 5)), choose a number of pixels and calculate the average value of those pixels. The two algorithms differ because they select different pixel groups for averaging. In a block averaging algorithm, the image is divided up into blocks of pixels, and the average is computed from the set of pixels in each block. In the example below, the blocks consist of 9 pixels in a 3 x 3 grid - the average is the sum of those nine pixels divided by 9. The '381 patent discloses that each of the pixels in a block share the same local average, meaning that each pixel is only involved in a single average calculation. (J.A. at 9, ('381, col. 3,1. 67 - col. 4, 1.7).) Block Averages
A sample block The next block

0 4D 0 4D 4D: 49 (D CD

6 M [GO so ^0

ID 6 4D

mmm^^mm^^ '^^m^oo 0 0 00 4) 06 mm m 0 0 : 0 ·m
The block average for pixel Isis: · + +...+ 9

The block average for pixel 0 and every other pixel in the block is: it + 0+4D +._+10

The same number is the block average for other pixel in the block

®, ® and every

9

6

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 12 of 46

Unlike a block average, a low-pass filter computes a new average for each pixel. The local average for a pixel is computed by averaging that pixel with its neighbors. The low-pass filter takes a selected pixel and all neighboring pixels, then computes the average of that group. It then repeats this operation for the next pixel, and so on for the entire image. A low-pass filter also differs from a block average because the pixels in a low-pass filter are included in multiple local averages: their own local average, plus the local averages of their neighbors. (J.A. at 9, ('381 col. 3,11. 61-67).) Low-Pass Filters
Low-Pass filter for pixel

Low-Pass filter for pixel

®

Low-Pass filter for pixel

s

t^ !fl t^! t^ !fit to tai ^ <#>
The Low-Pass fliter for pixel
The Low-Pass fliter for pixel 0is: · + is +.,,+ it 9

49

is:

+ 4V +... + it
9

The Low-Pass fliter for pixel Ois: f^+ E +.-+ 4D
9

Every other pixel has a different Low-Pass filter.

The above example indicates how the average of a given pixel is included in multiple local averages, when the average is calculated using a low-pass filter. As illustrated above, pixels 2,4 and 3,4 are included in the local averages associated to both pixel 2,3 and pixel 3,5. B. The '381 Patent and Its Prosecution History. 1. The '381 Patent.

The heart of Dr. Song's and Dr. Levinstone's idea is using a local averager and specific ratio algorithm to pick gamma functions to enhance an image. Three steps are involved. First,

7

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 13 of 46

an averager calculates the average luminance value for a set of pixel luminance values. Second, a gamma determining circuit uses the average value calculated in a ratio algorithm to set a gamma value. Third, a transfer function imposing circuit uses the gamma to transform the luminance value of each input pixel. Of particular importance is that the same data that is averaged is also transformed. There is no disclosure of transforming signals other than those that were averaged. a. Calculating a Local Average Luminance

Before the average can be calculated, a luminance value must be extracted from the color image for each pixel. For example, RGB Data captured by the image sensors is immediately subject to gain control and the luminance ("Y") is calculated according to the formula Y = 0.39R + 0.59G + 0.11B. (J.A. at 9, ('381, col. 3,11.40-56).) Averaging and transforming are performed only on this luminance "Y" portion of the image pixels. (See J.A. at 6-7 ('381, Figs. 1 & 4).) There is no disclosure of what, if anything, is done to the two "chrominance" portions of the image for particular pixels.4 These luminance values are then averaged using a block average. As is discussed in more detail below, although a low-pass filter is disclosed in the specification, the claims were written to describe only the use of a block average. b. Choosing a Gamma Function

The average is then used to choose a gamma function for the pixels involved in the average. (J.A. at 9 ('381, col. 4,11. 26-33).) As noted above, a gamma function increases contrast in the light areas when the gamma exponent is below 1, and increases contrast in the dark areas when the gamma exponent is above 1. In the '381 patent, the average value (a

4 In fact, the '381 patent does not even disclose structure for computing the two chrominance values for pixels in the image, nor any use for the chrominance signals. 8

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 14 of 46

number between 0 and 255) is divided by 128 minus 1, producing a value between 0 and 2.5 (Id. ('381, col. 4,11. 26-39).) The result of this calculation is then used, in part, to determine the value of gamma. Id. This method of converting a number between 0 and 255 (i.e. the average value) into a number between 0 and 2 is important because this "ratio" limitation was added to overcome the USPTO's obviousness rejection, as described below. Consequently, the calculation of this ratio is the central limitation of the '381 patent. c. Applying the Gamma Transfer Function

Lastly, the gamma function is applied to relevant pixel(s). (J.A. at 9 ('381, col. 4,11. 5662.) In dark areas, the contrast of dark pixels is enhanced. Similarly, in light areas, the contrast of light pixels is enhanced. Consequently, the overall contrast in various regions of the image is improved. 2. The Prosecution History

The prosecution history of the '381 patent is brief. In fact, there was only a single office action prior to the examiner allowing the claims of the patent to issue. The rejections of the examiner and the patentee's response to overcome them, however, are critical to understanding the proper scope of the invention claimed by the '3 81 patent. The examiner rejected every claim of the original application. First, the examiner found the dependent claims indefinite because several terms lacked an antecedent basis. (J.A. at 48.) Specifically, the examiner found that there was no antecedent basis for seven terms found in the dependent claims, including "the ratio", "the value", "the dynamic range" "the logarithm", "the antilogarithm", "the maximum value", and "the amount." Second, the examiner found the independent claims obvious over Okada, U.S. Patent No. 4,489,349. (J.A. at 49.) The examiner

5 This number is subsequently massaged to typically produce a value between 1/2 and 2. 9

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 15 of 46

found that Okada rendered the filed claims obvious because "[b]oth systems show an averaging circuit and a correction circuit which used the averaged information to produce an output which follows the slopes of the curves shown in Figure 2 of the present invention and Figure 2 of Okada."

F1 G. 2
FIG 2

INPUT
(Compare Fig. 2 of U.S. Patent 4,489,349 (J.A. at 88 (Okada)) with Fig. 2 of the '381 patent (J.A. at 6).) In addition, the examiner found that Okada rendered the filed claims obvious as Okada disclosed a "brightness control circuit having an average picture level detector 20 which averages input picture information and provides a control signal to a variable correction circuit 10." (J.A. at 49.) This variable correction circuit, in turn, "operates on the input-output signal to vary the characteristic of the input-output signal as a function of the detected average picture level detector (see Fig. 2)." (Id.) Finally, Okada disclosed the control of "relative brightness of the video signal such that the picture areas containing most of the picture information are corrected to give greater contrast." (Id.)

10

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 16 of 46

In its response to this office action, the patentee added the phrase "each signal having a value within a determinate dynamic range of values" to the preambles of claims 1 and 7.6 (J.A. at 53 & 56-57, respectively). The patentee explained that these amendments were necessary to "obviate[e] the Examiner's rejection based on a lack of antecedent basis for these terms." (J.A. at 58-59.) Also, in response to the examiner's obviousness rejections, the patentee further amended the original independent claims to include an additional limitation that included the following phrase: "wherein said transfer function is selected further as a function of the ratio of the value of the average electronic information signal to a select proportionate value of the dynamic range...". (See J.A. at 54 & 57.) This amendment was necessary to overcome the obviousness rejection in light of Okada. Without the language relating to the ratio, the patentee was merely claiming the general slope of the function described by Okada, a claim that was rejected by the examiner. II. THE LAW OF CLAIM CONSTRUCTION A. Generally

In claim construction, "[t]he construction that stays true to the claim language and most naturally aligns with the patent's description of the invention will be, in the end, the correct construction." Phillips v. AWH Corp., 415 F.3d 1303, 1316 (Fed. Cir. 2005) (en banc). As such, the specification "is the single best guide to the meaning of a disputed term." Id. at 1315 (internal quotations and citations omitted). The importance of the specification derives from the statutory mandate that an inventor "provide a `full' and `exact' description of the claimed invention." Id. at 1316. Nevertheless, the specification is not the only source that provides guidance in the course of claim construction. Like the specification, the prosecution history "inform[s] the meaning of the claim language by demonstrating how the inventor understood the

6 The claim numbers refer to the claims as issued in the '381 patent. 11

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 17 of 46

invention and whether the inventor limited the invention in the course of prosecution, making the claim construction narrower than it otherwise would be." Id. at 1317. Additionally, the court may rely on extrinsic evidence but it should be "considered in the context of the intrinsic evidence." Id. at 1319. B. Means-Plus-Function Claims

The '381 patent includes means-plus-function claim language. Section 112,16 states that means-plus-function limitations "shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof." 35 U.S.C. § 11216. The presence of the word "means" in a claim term creates a rebuttable presumption that the limitation should be construed under 35 U.S.C. § 112 16. Cole v. Kimberly-Clark Corp., 102 F.3d 524, 530-31 (Fed. Cir. 1996). In construing means-plus function limitations, the Court must first identify the claimed function, and then identify the corresponding structure in the specification. Biomedino, LLC v. Waters Techs. Corp., 490 F.3d 946, 950 (Fed. Cir. 2007). But, the "structure disclosed in the specification is `corresponding' structure only if the specification or prosecution history links or associates that structure to its function." B. Braun Med., Inc. v. Abbott Labs., 124 F.3d 1419, 1424 (Fed. Cir. 1997). C. Preamble

A preamble is limiting on the claim if "the claim cannot be read independently of the preamble and the preamble must be read to give meaning to the claim or is essential to point out the invention ." Porter v. Farmers Supply Serv., Inc., 790 F.2d 882, 885, (Fed. Cir. 1986). In particular, a preamble is limiting when it contains definitions that serve as an antecedent basis for terms used in the body of the claims . Electro Scientific Indus., Inc. v. Dynamic Details, Inc, 307 F.3d 1343, 1348 (Fed. Cir. 2002).

12

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 18 of 46

III.

HEWLETT-PACKARD' S PROPOSED CONSTRUCTIONS The parties have asked the Court to construe a number of terms relating to claims 1 and 7

of the '381 patent-the only independent claims asserted in this litigation-as well as dependent claims 2-3 and 8-9. The number of terms at issue in this case is dictated by the form that the patentee elected to claim his invention. While claim 7 is a method claim, claim 1 is a meansplus-function claim that requires the Court to construe two separate functions as well as their corresponding structures. A. The Preamble And Its Terms Are Limiting

Prior to undertaking the construction of the claims, the Court must determine whether the preambles of independent claims 1 and 7 act as limitations of the claims. Hewlett-Packard contends that applicable law requires a finding that the preambles limit the claims. First, the preambles of independent claims 1 and 7 define a number of terms that are referred to later in the claim, including "electronic information signals" and "dynamic range". The law is settled that a preamble acts as a claim limitation when it provides an antecedent basis for terms used in the body of the claim. Electro Scientific Indus., 307 F.3d at 1348. For both of these terms, the inclusion in the preamble was necessary to provide an antecedent basis and they are properly considered limitations of the claim. For other terms, the specification makes clear that the terms are necessary to breathe life and meaning into the claims, including "continuously enhancing", "electronic image data received in a continuous stream of electronic information signals" and "each signal having a value within a determinate dynamic range of values." See Poly-Am., L.P. v. GSE Lining Tech.,

Inc., 383 F.3d 1303, 1310 (Fed. Cir. 2004) (noting that "when reciting additional structure or steps underscored as important by the specification, the preamble may operate as a claim

13

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 19 of 46

limitation") (quoting Catalina Mktg. Int'l, Inc. v. Coolsavings.com, Inc., 289 F.3d 801, 808 (Fed. Cir. 2002)). As each of the preamble terms for which Hewlett-Packard seeks construction provides crucial context to, and is necessary to explain the meaning of the claims, the Court should construe the terms as set forth below. B. "Continuously enhancing...." (claims 1,7) HP Proposed Constr. successively transforming Polaroid Proposed Constr. successively transforming

Claim Term A system/method for continuously enhancing

Support for HP Position : Preamble limitation that describes how the "transforming" means/ step is accomplished. Why Polaroid Position is Incorrect : The preamble provides further meaning to the term "transforming" that is in the body of the claim. Polaroid and Hewlett-Packard agree on the construction of "continuously enhancing." The only dispute regarding this term is whether it should be construed at all. Polaroid maintains that the term does not need to be construed because it is found in the preamble. The term "continuously enhancing," however, is not only found in the preamble, but is used throughout the patent. (See J.A. at 5 ('381 at Abs.) (defining the claimed invention as "[a] system and method ... for continuously enhancing electronic image data"); J.A. at 8 ('381, col. 2, 11. 63-65) (noting that "the system and method of this invention operates to continuously enhance electronic image data"); see also J.A. at 10 ('381 col. 5,11. 8-15) (noting that "gamma y changes continuously .... so that each image defining luminance electronic information signal is enhanced ...."). The specification is, in fact, replete with references to the importance of "continuously enhancing" and, as such, the term cannot be ignored when it appears in the preamble. See Poly-Am., L.P. v. GSE Lining Tech., Inc., 383 F.3d 1303, 1310 (Fed. Cir. 2004) (noting that "when reciting additional structure or steps underscored as important by the

14

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 20 of 46

specification, the preamble may operate as a claim limitation") (quoting Catalina Mktg. Int'l, Inc. v. Coolsavings.com, Inc., 289 F.3d 801, 808 (Fed. Cir. 2002)). Further, the term is limiting because it provides information as to how the "transformation means" (claim 1) and "transforming" method (claim 7) are accomplished. See Porter v. Farmers Supply Serv., Inc., 790 F.2d 882, 885, (Fed. Cir. 1986) (A preamble is limiting if "the claim cannot be read independently of the preamble and the preamble must be read to give meaning to the claim or is essential to point out the invention."); see also Biacore v. Thermo Bioanalysis Corp., 79 F. Supp. 2d 422, 457 (D. Del. 1999) (Robinson, J.) (finding the preamble limiting where the phrase in the preamble "informs the construction of the remainder of the claims"). The term "continuous" serves to provide a further limitation in that the "transforming" is continuous - that is successively transforming. The patent states: "[t]oward that end, the system and method of this invention operates to continuously enhance the electronic image signal received in a continuous stream of electronic information signals...." (J.A. at 8 ('381, col. 2,11. 62-65).) Dictionaries commonly define continuous as "uninterrupted in time, sequence or essence; going on without interruption", "without interruption of substance" and HewlettPackard offers the same construction. (The Oxford English Dictionary, ("OED"), Ex. E at 830). The combination of "continuously enhancing" with a transforming step (claim 7) or a transforming means (claim 1) indicates how the image is enhanced. Because this term is critical to the invention described in the specification and must modify the term "transforming" found later in the claims, this claim term should be construed. See Biacore, 79 F. Supp. 2d at 457. C. "Electronic Information Signals" and "Electronic Image Data Received in a Continuous Stream" (Claims 1 and 7)

"Electronic information signals" is a term that is found in the preamble, and the parties do not dispute that this term requires construction. Nevertheless, the parties disagree on three

15

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 21 of 46

meaningful points in their constructions of "electronic information signals" (claim 7) and "electronic image data received in a continuous stream of electronic signals." (claim 1). First, Hewlett-Packard proposes a construction limiting electronic information signals to luminance data, consistent with the only disclosed usage. Polaroid seeks an unsupported but broader construction, including color and chrominance values, which are not described as part of the invention in the specification. Second, Hewlett-Packard construes the claim so the "continuous" stream is "uninterrupted", whereas Polaroid ignores the ordinary meaning of the term and, instead, argues the term merely means a "successive series of signals". Third, Hewlett-Packard argues that the image data must constitute "an original image to be recorded" in contrast to Polaroid's mere "pixel information" construction. 1. "Electronic Information Signals" Mean Luminance Signals.

Polaroid Proposed Constr. HP Proposed Constr. Claim Term signals providing pixel information, such signal(s) providing luminance electronic as color, luminance, or chrominance values pixel information information signals Support for HP Position : The '381 disclosure indicates that Luminance (Y) is the only electronic information signal which is ever transformed and averaged, each of which is recited later in the claim. Why Polaroid Position is Incorrect: Broadly encompasses "color" and "chrominance" values which is beyond the disclosure of the patent. According to the claims at issue in this litigation, the "electronic information signals" are subject to an "averaging means" (claim 1) or the step of "averaging" (claim 7), as well as a "transforming means" (claim 1) or the step of "transforming" (claim 7). The only type of signals disclosed in the patent specification that are subject to the claimed steps and means are lumenance signals. (See generally, e.g., J.A. at 9 ('381, col. 4,1. 56-col. 5,1. 3) (noting that Yout is the imaging defining luminance electronic information signal transformed in the manner of this invention by "transfer function imposing circuit 16").) This is made clear in Figures 1, 2, and 4, which show luminance "Y" being the input to averager 12 and being transformed in box 16. 16

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 22 of 46

CONTROL PARAMETER

FIG 1
''INPUT

C

Z7
AVERAGER AV

14

y=(1+C)

(Av/M 1)

POUT ° YMAX (YIN /YMAX )

i

0.

''OUTPUT

The claimed "electronic information signals" are signals representing the luminance of the image, received from an input device such as a CCD array. As the specification makes clear: The electronic information signal values retrieved from the photosensitive array in this manner are preferably converted to luminance (Y) and chrominance, e.g., (R-Y and B-Y) signal values. For the case where the two-dimensional photosensitive array is overlayed with red, green and blue filters, the luminance electronic information signals are preferably determined by the following relationship : Y=0.30R+0.59G+0.11B as is well known in the television art (J.A. at 9 ('381, col. 3, 11. 35-43) (emphasis added).) The value of gamma is thereafter directed to a transfer function imposing circuit 16 which operates to impose the following transfer function on the image defining luminance electronic information signals (Y) .... (J.A. at 9 ('381, col. 4,11. 56-59) (emphasis added).) Your equals the image defining luminance electronic information signal transformed in the manner of this invention to provide an enhanced image. (J.A. at 9 ('381, col. 4, 1. 67-col. 5, 1. 3).) As stated above, the patent specification, including Figures 1, 2 and 4, follow the well known convention of representing "luminance" using the variable "Y." (J.A. at 9 ('381, col. 3,

17

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 23 of 46

REDACTED
1. 37).

Consequently, the term should be construed in accordance with its use in the specification. See Phillips, 415 F.3d at 1315. Polaroid's construction of "electronic information signals" is contrary to the specification, because it allows for "color, luminance, or chrominance values." (J.A. at 9 ('381, col. 3, U. 59-60) ("The image defining luminance electronic information signals...").) The specification does not disclose either averaging or transforming color (R,G,B) or chrominance values (Cl, C2), and Polaroid should be precluded from urging a construction that encompasses an invention that is neither enabled nor described in the written description. LizardTech, Inc. v. Earth Resource Mapping, Inc., 424 F.3d 1336, 1345-46 (Fed. Cir. 2005).

7 The Federal Circuit has relied on inventor testimony to determine what is claimed by the patented invention. See Automotive Techs. Int'l, Inc. v. BMW of N. Am., Inc., 501 F.3d 1274, 1283 (Fed. Cir. 2007) (relying on testimony of co -inventor in interpreting the specification). 18

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 24 of 46

2.

Electronic image data received in a continuous stream of electronic information signals

Polaroid Proposed Constr. Claim Term HP Proposed Constr. an uninterrupted stream of electronic data received in a electronic image data successive series of signals providing received in a continuous received luminance image pixel information, such as color, data [pixels] defining an stream of electronic on inal image to be recorded luminance, or chrominance values information signals The plain meaning of "continuous" means not interrupted, and the Support for HP Position : specification indicates that the processing is of an original image to be recorded. Why Polaroid Position is Incorrect: Broadly encompasses "color" and "chrominance" values which is beyond the disclosure of the patent; not limited to original images to be recorded. The construction of the term "electronic image data received in a continuous stream of electronic information signals" can be divided into two distinct disputes between the parties. First, the parties dispute the meaning of "continuous." Second, the parties dispute how the "electronic image data" is "received." a. Hewlett-Packard's Proposed Construction of "Continuous" Properly Relies on the Plain and Ordinary Meaning of Continuous.

In the context of a stream of signals, the plain meaning of continuous is "uninterrupted." (OED, Ex. E at 830). As such, Hewlett-Packard has adopted this plain meaning as part of its construction for the phrase "electronic image data received in a continuous stream of electronic information signals." Polaroid's proposed "successive" construction on the other hand, fails to add any meaning, as everything in a "stream" would already be successive. A "continuous" stream is a stream where there is no interruption to the data within the stream. b. Hewlett-Packard 's Proposed Construction Regarding How the "Electronic Image Data is Received " Comports with the Specification.

Polaroid's proposed construction also fails because it attempts to broaden the scope of the patent claims to include subject matter that was not disclosed by the '381 patent. The Court should reject Polaroid's improper attempt to broaden the scope of the claimed invention through claim construction. See Netword, LLC v. Centraal Corp., 242 F.3d 1347, 1352 (Fed. Cir. 2001)

19

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 25 of 46

REDACTED
("The role [of claim construction] is neither to limit nor to broaden the claims, but to define, as a matter of law, the invention that has been patented.") The specification indicates that the input luminance signals are "received" from an imaging device that has captured data relating to a scene in the real world, such as a CCD array that captures image data in a digital camera. For example, in the Summary of the Invention, the patentee states that his invention is: A system is provided for enhancing electronic image data received in a continuous stream of electronic information signals wherein each signal corresponds to one of a plurality of succeeding pixels. The pixels collectively define an image to be recorded. (J.A. at 8 ('381, col. 1, 1. 65-col. 2,1. 5) (emphasis added).) Later, in the specification, the patentee reiterates that the invention is one that receives data from an imaging device, as follows: The electronic information signals input at terminal Yinput may be derived in a well-known manner by a two-dimensional photosensitive array or sensor (not shown) which may comprise a high resolution charge coupled device (CCD) or charge injection device (CID). The sensor receives image scene light in any well-known manner by way of an objective lens and shutter (also not shown). (J.A. at 9 ('381, col. 3, 11. 6-13).)

Toward that end, the specification citations above indicate that the system and method of this invention operates to receive an original image from an originally sensed scene that is to be

20

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 26 of 46

recorded, as opposed to an existing image on a computer hard drive that has already been recorded. Again, the specification makes clear that this is the only reasonable interpretation when it describes the nature of the problem to be solved by the claimed invention. For example, the specification states: Difficulties arise, however, as a result of differences between the wide dynamic range of the scene originally sensed and recorded and the substantially smaller dynamic range to which a photographic print may be exposed. The wide dynamic range of luminance intensities within the scene originally recorded may thus be compressed or clipped to the substantially smaller dynamic range of the photographic print, losing detail within certain portions of the dynamic range that were otherwise visible in the original scene. Thus, it may be desirable to transform the original image defining electronic information signals in a nonlinear manner to selectively increase and/or decrease the contrast and brightness in certain portions of the scene such as those that might be brightly lit by sunlight or underlit as a result of shadows. (J.A. at 8 ('381, col. 1,11.26-40) (emphasis added).) This confirms that the inventors wanted to alter the image as originally sensed from the real world prior to recording them in the digital imaging device. Polaroid's "pixel information" construction is incorrect because it does not indicate that the signals represent an image to be recorded, perhaps indicating that an already-recorded and processed image would fall within the scope of its construction. This is also not supported by the specification or written description as it attempts to expand the invention beyond a camera or other image capture device, the application for which Polaroid created the invention. Polaroid's construction does not even include the idea of an image, effectively vitiating the entire limitation. The construction cannot be correct. D. "Each signal having a value within a determinate dynamic range of values" (claims 1,7) Polaroid Proposed Constr. each signal being associated with a value that lies within a range of possible values bounded by definite limits

HP Proposed Constr. Claims Claim Term each received pixel has an each signal having 1, 7 associated luminance value that a value within a lies within a predetermined determinate group of luminance values dynamic range of values Support for HP Position : Supported by specification 21

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 27 of 46

Why Polaroid Position is Incorrect : Leads to infinite possibilities and is not disclosed. This preamble term is limiting because "dynamic range" is recited further in the body of claims 1 and 7 and thus the use of the term in the preamble provides the antecedent basis for the claim term as described above. See Bicon, Inc. v. Straumann Co., 441 F.3d 945, 952-53 (Fed. Cir. 2006) (finding the preamble limiting because the terms in the body of the claim "derive their antecedent basis from the preamble); see also ABB Automation Inc. v. Schlu)nberger Res. Mgmt. Servs., Inc., 254 F. Supp. 2d 475, 477 (D. Del. 2003) (Robinson, J.) (finding that, where term in preamble provided antecedent basis for term later in claim, argument that preamble was not limiting "has no merit"). Moreover, the prosecution history also supports this conclusion. The original claims were rejected by the patent examiner as being indefinite under 35 U.S.C. § 11212 due to a lack of an antecedent basis for a number of claim terms. (J.A. at 48.) In response to this rejection, Polaroid amended the preamble to provide the necessary antecedent basis noting that "[c]laims 1, 7 and 8 have also been amended to recite that the signals have values within a determinate dynamic range thereby obviating the Examiner's rejections based on lack of antecedent basis for these terms." (J.A. at 58-59.) Not only does the preamble define, and therefore limit, but Polaroid should be estopped from arguing to the contrary in light of its actions during prosecution. The principal difference between the parties' constructions is that Hewlett-Packard's construction limits the range of possible pixel values to a predetermined group. In the 8-bit context, the group of values would consist of the integers ranging from 0 to 255, inclusive for 256 distinct values. Polaroid's construction permits the pixel to take any value within its boundaries. In the 8-bit context, Polaroid's pixels could be non-integers or any other "possible value" between 0 and 255. Polaroid's construction would lead to an absurd result, an unlimited

22

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 28 of 46

dynamic range, because there are an infinite number of possible fractional numbers between 0 and 255. The patent' s plain language indicates that the luminance signals8 must have a value that is within a "determinate" (known) specific number of values. This indicates that the overall range of values is a known set that cannot be departed from. The specification so states: The analog luminance electronic information signal values for each pixel element of the photosensitive array for the example herein described are digitized to an 8-bit binary number so as to have a dynamic integer range of from 0 - 255 within which range are 256 intensity levels and a maximum luminance value of YMAX =255. (J.A. at 9 ('381, col. 3,11.43-51).) In light of this clear statement in the specification , Polaroid' s construction is incorrect because it defines a value within the dynamic range to be "a value that lies within a range of possible values " which would allow for values outside the dynamic range to be valid. For example, the patent states that for an 8-bit system there are 256 values. The lowest value is 0, the next lowest is 1, and the maximum value is 255. Polaroid 's construction would appear to allow a value of 250.5 to be within the dynamic range, even though it is impossible to represent 250.5 in an 8-bit system and even though it would actually provide an infinite number of values. Polaroid' s construction also ignores the very purpose of the claimed invention, namely to compensate for the limited number of values in a digital system as opposed to the infinite range of values available in the real world. E. 1. "Averaging" Claim Terms "Averaging" (claims 1, 7) / "Average" (claims 1, 2, 7, 8)

8

The term " each signal" refers to the "electronic information signals" and the only signal that is disclosed is the luminance signal : "YMAx equals the highest value of the dynamic range for the electronic information signals or 255 for the example herein described. Y,,ut equals the image defining luminance electronic information signal transformed in the manner of this invention to provide an enhanced image." (J.A. at 9-10 ('381, col. 4,1. 66 - col. 5,1. 3).) 23

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 29 of 46

Polaroid Proposed Constr. calculating an intermediate value for of calculated intermediate value signal providing pixel information, such as a color, luminance, or chrominance value of calculated intermediate value Support for HP Position : Plain language and ordinary meaning. Claim Term averaging average average electronic information signal Why Polaroid Position is Incorrect : Polaroid's construction would encompass almost any mathematical operation, including those unrelated to averaging. Within the '381 patent's specification, there is no definition of "average" or "averaging" other than a block average (average of a block) or low pass filter average (pixel-by-pixel average). The term "average" is not defined and should be given its plain meaning. Using its plain meaning, an "average" of a set of numbers is determined by adding up all the numbers and dividing the result by the number of numbers you added up. This is an "arithmetic mean." (See Lial and Homby, Intermediate Algebra, Ex. F at 701 ("The arithmetic mean, or average, of a group of numbers... is found by dividing the sum of the numbers by the number of numbers.").) This is a case where the ordinary meaning of claim language is readily apparent "and claim construction requires little more than the application of the widely accepted meaning of commonly understood words." See Phillips, 415 F.3d at 1314. Polaroid's construction of "an intermediate value" has no support in the specification and is broad enough to include operations that do not involve the commonly understood meaning of averaging. For example, the median of a group of numbers is an "intermediate value", but it is not an average; the square root of a sum of numbers is an intermediate value, but not an average. Polaroid's construction essentially erases the term "average" from the claim in favor of nearly any mathematical operation and is, therefore, wrong. 2. Claim Term means for averaging "Means for Averaging ... and Providing an average" (claim 1) nstr HP Proposed , avera e Function : providing an 24 Polaroid Proposed Constr. Function : averaging electronic

HP Proposed Constr. taking an arithmetic mean of an arithmetic mean No construction necessary. Alternatively: the average of the electronic information signals.

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 30 of 46

electronic information signals corresponding to selected pluralities of pixels and providing an average electronic information signal for each said plurality of pixels so averaged

for selected pixel values around one pixel, where the average is correlated to each pixel making up the average.

Disclosed Structure : a block averager 12 with a buffer memory that takes luminance as an input Disclosed Structure : a low pass and outputs an average luminance filter or block average and value that is correlated to each pixel in the block, and equivalents equivalents thereof thereof. Support for HP Position : Averaging disclosed col. 3:61- col. 4:25. Only one type of average, the "block" average, has an average associated with each pixel in the plurality of pixels averaged. A buffer memory is always required (col. 4,11. 8 -9). The disclosed averager operates on the luminance. Figs. 2, 4 & col. 4,11. 25-30. Why Polaroid Position is Incorrect: The claim language excludes the "low pass filter" embodiment and Polaroid's construction of "averaging" as "calculating an intermediate value for" erases the "avera in " limitations. The parties agree that this phrase is subject to a "means plus function" claim construction

information signals corresponding to selected pluralities of pixels and providing an average electronic information signal for each said plurality of pixels so averaged.

pursuant to 35 U.S.C. 112 16. Accordingly, the Court must evaluate the claimed function and construe the claim as limited to the disclosed structure that clearly performs the identified function and equivalents. Cardiac Pacemakers, Inc. v. St. Jude Med., Inc., 296 F. 3d 1106, 1113 (Fed. Cir 2002) ("In order to qualify as corresponding, the structure must not only perform the claimed function, but the specification must clearly associate the structure with performance of the function." (emphasis added)) The function is precisely stated: averaging electronic information signals corresponding to selected pluralities of pixels and providing an average electronic information signal for each said plurality of pixels so averaged; (J.A. at 11 ('381, claim 1) (emphasis added).) Like the function of the asserted claim, the means are clear in the specification. There are only two methods of "averaging" disclosed in the '381 patent's specification- a "block average" and a "low-pass filter." However, although both a "block average" and the calculation of an

25

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 31 of 46

average through the use of a low-pass filter are disclosed in the specification , only a block average can be used in light of the entire claim. In a block average, as noted above, each pixel is only included in the average of its block. This makes block averaging computationally easier, since averaging happens one time a group of pixels inside the block . (J.A. at 9 ('381 , col. 4,11. 15-20).) For a block average of 4x4 pixel sized blocks , one average is associated with all 16 pixels . A low-pass filter uses a moving window, so a new average must be computed for every single pixel, which requires additional computational capacity as compared to a block average, but is more accurate . (See id.) For a 4x4 pixel low -pass filter, 16 averages are computed. The claim language recites "an" "average electronic information signal for each said plurality of pixels so averaged" and must be read in the context of the entire claim . Phillips, 415 F.3d at 1314 ("[T]he context in which a term is used in the asserted claim can be highly instructive"). Later in the claim , the claim language requires that a transfer function "is selected as a function of the electronic information signal for one pixel and the average electronic information signal for the select plurality of pixels containing said one pixel ." Taken in combination, this language requires that each pixel be contained in only one of the "select plurality of pixels" averaged by the system . A low-pass filter does not have this property, and therefore is not encompassed by the claim language and the structure should be limited to a block average. . In addition to the block average, the structure must also include a buffer memory because, in order to average a group of values, one needs a device that can hold the values as they arrive so that the performed function can be performed on this stored set of values. The specification supports this additional structure : "Both low pass filtering and block averaging require a buffer memory to hold the selected groupings of pixel values...." (J.A. at 9 ('381, col. 4, 11. 8 - 10 (emphasis added)).)

26

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 32 of 46

REDACTED
Finally, the structure must be a structure that receives luminance values and averages these values. In the specification, the only signals shown entering averager 12 are Y (luminance) signals, and there is no disclosure of averaging any other kind of signals. (See generally J.A. at 6 ('381, Fig. 1).)

Given the specification and the testimony of the inventor, it is clear that luminance signals are the only signals that are averaged by the averager 12. (J.A . at 6 ('381 , Fig. 1); see J.A. at 9 ('381 , col. 3.,11.59-61 ) (" The image defining luminance electronic information signals

27

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 33 of 46

are thereafter averaged ... by an averager 12".).) It is improper to construe a claim to cover embodiments that are not disclosed, and allowing any signals other than luminance signals in averager 12 would violate this rule. LizardTech, 424 F.3d at 1345-46. For these reasons, Hewlett-Packard's proposed construction should be adopted by the Court. F. "Selecting" (claim 7)

Claim Term Polaroid Pro posed Constr. HP Proposed Constr. selecting one of a plurality of Selecting one of a plurality each input pixel has an different transfer functions for the of different transfer associated transfer functions for the electronic function out of different signal providing pixel information, such as a color, information signal for each transfer functions , and luminance, or chrominance of the plurality of the transfer function is value for each of the plurality of succeeding pixels in a selected based on the manner whereby each input pixel value, and the succeeding pixels in a manner transfer function is selected average that was formed whereby each transfer function is selected as a function of the signal using the input pixel as a function of the providing pixel information, value, where each input electronic information such as a color, luminance, or signal for one pixel and the pixel is part of only one chrominance value for one pixel average. average electronic and the calculated intermediate information signal for the value for the select plurality of select plurality of pixels pixels containing said one pixel containing said one pixel. . The plain language of the claim and specification. Support for HP Position : Why Polaroid Position is Incorrect: Polaroid's construction is inconsistent with the specification. The construction of the claim 7 step of "selecting" should be governed by the plain meaning of the words used in the claim language. The phrase "selecting one of a plurality of different transfer functions for the electronic information signal for each of the plurality of succeeding pixels...." means that each input pixel has an associated transfer function out of different transfer functions. This is shown in Figure 2, which identifies a number of transfer functions based on different values of gamma. The next phrase "in a manner whereby each transfer function is selected as a function of the electronic information signal for one pixel and the average electronic information signal for the select plurality of pixels containing the one pixel" means that the transfer function is selected

28

Case 1:06-cv-00738-SLR

Document 110

Filed 01/18/2008

Page 34 of 46

based on the input pixel value and the average that was formed using the input pixel value. This has to be an average of "x" values, one of which is the input pixel. A low pass filter cannot perform this claimed averaging because a low pass filter would require that each input pixel value is part of multiple averages. This would cause the language "the average electronic information signal for the select plurality of pixels containing said one pixel" to be violated, and the construction would not be correct. To avoid such a result, HewlettPackard includes "where each input pixel is part of only one average" in the construction which is supported by the disclosed "block" average. See Phillips, 415 F.3d at 1314 ("[T]he context in which a term is used in the asserted claim can be highly instructive"). Polaroid's construction contradicts the plain language and the specification in that it allows for the pixel information to be a "color, luminance, or chrominance value for one pixel" where the specification discloses transfer functions acting on luminance, not color or chrominance. It also substitutes "calculated intermediate value" for the word "averaging" in the claim, which is inconsistent with the words of the claim and completely unsupported in the specification and should be rejected. G. "A select proportionate value of the dynamic range of the electronic information signals" (claim 7) HP Proposed Constr. any value within the determinate dynamic range of values, selected depending on where the least image enhancement is desired Polaroid Proposed Constr. value within the range of possible values

Claim Term a select proportionate value of the dynamic range of the electronic information signals

Support for HP Position : The select proportionate value is defined in the specification. Why Polaroid Position is Incorrect : Polaroid's construction is inconsistent with the specification. The parties agree that "a select proportion