Case No. Heisei 7 (Wa) 13135
Case Of Demand For Injunction of Manufacture or The Like Based upon The Patent Right
translated by Takeo Hayakawa
JUDGMENT

Plaintiff:  Ikegami Patent Incubator, Ltd.
   No. 12-25, Tachibana-cho 2-chome
   Amagasaki-shi, Hyogo-ken
Representative Director: Kimiko IKEGAMI
Attorneys-at-law representing the Plaintiff
   Ryuichi MURABAYASHI
   Tsukasa MATSUMOTO
   Kouji MATSUMOTO

Defendant:  Sharp Corp.
   No. 22-22, Nagaike-cho, Abeno-ku
   Osaka-shi
Representative Director: Haruo TSUJI
Attorneys-at-law representing the Defendant    Keizo KOUSAKA
   Nakaroku TORIYAMA
   Yasuaki IWAMOTO
Patent attorneys assisting the above-identified
attorneys-at-law:
   Hisao FUKAMI
   Hachiro MORISHITA

Text of Judgement

 1. All of the Plaintiff's claims shall be dismissed.
 2. Litigation expenses shall be borne by the Plaintiff.

Facts and Reasons

I.  Claim

 1. The Defendant must not manufacture a device identified in a list of the accused device (submitted by the Plaintiff).
 2. The Defendant shall pay to the Plaintiff money in the amount of 75,000,000 yen as well as money accruing therefrom at an annual interest rate of 5 % during a period starting from December 28, 1995 up to a date when the payment will be completed.

II.  Gist of Facts of the Case

I.  Facts Forming a Basis (Verification of whether each of facts is a non-contestable fact based upon means of proof or the total gist of the arguments which will be presented hereinunder.)

1.  The patent right possessed by the Plaintiff
 The Plaintiff possesses the following patent right (hereinafter referred to as "the present patent right".)
 (i)  Title of the Invention:
  Information Processing Device
 (ii)  Filing Date: December 13, 1980
(Japanese Patent Application No. Sho 63-6595)
The present patent application is a divisional application from Japanese Patent Application No. Sho 55-176296 (hereinafter referred to as "the parent application".).
 (iii)  Publication Date: February 15, 1990
(Japanese Patent Application Publication No. Hei 2-7107)
 (iv) Registration Date: October 22, 1990
 (v) Japanese Patent No. 1583177
 (vi) Scope of Claim for a Patent (Claim 1)
 Claim 1 in the scope of claim for a patent of the specification (hereinafter referred to as "the present specification") attached to the application request of the patent application from which the present patent right arose is identified as what is claimed in Claim 1 of the patent publication ("the present patent publication") attached to this Judgement.  (The invention defined in claim 1 in the scope of claim for a patent of the present patent publication is patented and referred to as "the present invention").

2. Constituent Features of the Present Invention are Paraphrased as Follows:

 The Court finds that it is appropriate to paraphrase the constituent features of the present invention as follows:

E: An information processing apparatus comprising:
A: input means for inputting unprocessed information to be processed;
B: recognition means for recognizing the information inputted into said input means;
C: information processing means for, based upon the unprocessed information recognized by said recognition means, processing the information into processed information corresponding thereto; and
D: processing execution instruction means for causing said information processing means to execute a processing operation upon detection of terminator information which is present in the recognized information outputted from said recognition means.

3. Advantageous Effects of the Present Invention are found as follows:

(1) "According to the present invention, an information processing operation is executed upon detection of terminator information which is present in an output after recognized so that detecting accuracy can be considerably improved and auto-processing can be ensured, compared with a case where terminator information is detected from information before recognized including much noise."  (column 8, lines 17 to 22 of the present publication.)

(2)  "Generally, in a case where terminator information is detected from information before recognized, unless timing for instructing execution to the information processing means is delayed (e.g. by providing a delay unit) in consideration of a time delay (time lag) which is necessary for the recognition means to execute a recognition operation,
it is impossible to execute timely operation, or to obtain processed information with high accuracy as recognizing operation is still in mid-course at information recognition means notwithstanding the instruction for execution of operation is given from AS7* to recognition means.
In contrast, in the present invention, the recognized information becomes a target for detecting terminator information, and all the information to be processed is necessarily inputted in the information processing means every time the terminator information is detected.  Therefore, the present invention dispenses with the need to provide time controlling means such as the delay unit to execute the processing operation with a good timing and automatically.  (column 8, lines 22 to 36 of the present publication.)

4. Prosecution History of the Parent Application:

(1) Claim 1 of the specification first attached to a request of the parent application (hereinafter, referred to as "the original specification of the parent application") is identified as follows (Exhibit B2).
"An electronic translation apparatus characterized in that:
source language input unit is connected to a translation language output unit through an electronic translation unit, and a translation control unit is connected to said electronic translation unit, said translation control unit comprising means for detecting termination of a source language input signal and translation instructing means for causing the electronic translation unit to make a translating operation"

(2) Final decision of rejection was made on October 27, 1988 (Exhibit A21-2) in connection with the parent application, and thereafter, the final decision was concluded without any change.

5. Defendant*s Act in Question

 It is found that the Defendant has been manufactured and sold a Machine Translation System "DUET Qt" (hereinafter, referred to as "the accused device") since around December 1992.

II. Plaintiff's Demand

Subject Matter of This Case:
The Plaintiff claims injunction ‡@ and damages ‡A as against the Defendant on the ground that the accused device is encompassed within the technical scope of the present invention whereby the Defendant*s acts of manufacturing and selling such accused device are deemed to be an infringement of the present patent right:
‡@ injunction of manufacturing and selling the accused device based on the present patent right; and
‡A damages arising from an infringement of the present patent right.

III. Points at Issue

1. Constituent features of the accused device.
2. Whether or not the accused device is encompassed within the technical scope of the present invention.
3. A sum of compensatory damages payable to the Plaintiff in a case where the Defendant is proven to have an obligation to indemnify the Plaintiff for damages.
 

III.  Pleadings of Both Parties as to Points at Issue

I. Point at Issue 1
(as to the constituent features of the accused device)

[Plaintiff's Pleading]

1. The constituent features of the accused device should be identified as shown in List of the Accused Device submitted by the Plaintiff as attached hereto.
2. The constituent features of the accused device therein are considered as corresponding to those of the present invention, respectively as shown in List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Plaintiff as attached hereto.
3. The Defendant argues that operation for inserting a return code is not performed in Japanese language calling operation.  However, such return code insertion is actually performed in Japanese language calling operation.  Even if such argument made by the Defendant is supposed to be correct, the accused device comprises "operation for dividing a text into sentences one by one and calling them" in performing the Japanese language calling operation.

[Defendant's Pleading]

1. The constituent features should be identified as shown in List of Constituent Features of the Accused Device submitted by the Defendant as attached hereto.
2. The constituent features of the accused device therein when simplified for the sake of convenience are considered as corresponding to those of the present invention as shown in List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Defendant as attached hereto.
3. "Return code insertion" and "one sentence extract" in the accused device are performed in accordance with the following delimitation:
(1) Operation for inserting a "return code" after a specific character (a comma or a period) uniformly irrespective of whether or not it is one sentence in Japanese language registering operation;
(2) Operation for calling a sentence one by one which forms a translation unit based upon given conditions in Japanese language calling operation;
(3) Operation for inserting a "return code" based upon given conditions in English language registering operation; and
(4) The same operation as described in the above (3) in the case of English language calling operation.

II. Point at Issue 2
(as to whether or not the accused device is encompassed within the technical scope of the present invention)

[Plaintiff*s Pleading]

1. Presence of Literal Infringement:

 In view of Plaintiff's Pleading 2 concerning Point at Issue 1 (List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Plaintiff attached hereto), the accused device conforms to the constituent features of the present invention as follows:

(I) As to Constituent Feature A
(1) There is correspondency between "a series of plurality of sentences (text) including characters, punctuation marks, signs and codes" of Constituent Feature "a" of the accused device and "unprocessed information to be processed" of Constituent Feature "A" of the present invention.
(2) There is correspondency between "an image scanner 12" of Constituent Feature "a" of the accused device and "input means" of Constituent Feature "A" of the present invention.
(3) Hence, Constituent Feature "a" of the accused device conforms to Constituent Feature "A" of the present invention.

(II) As to Constituent Feature B
(1) There is correspondency between "a style of a character, a punctuation mark, a code or the like (image signal) read by the image scanner 12" of Constituent Feature "b" of the accused device and "information inputted into said input means" of Constituent Feature B of the present invention.
(2) There is correspondency between "by making comparison with characteristic data on a word (in English) or a character (in Japanese) which are prestored (resemblance calculation) whereby a character, a punctuation mark, a sign, or a code is recognized and then, converted into a character signal (character information)" of Constituent Feature "b" of the accused device and "recognize" of Constituent Feature B of the present invention.
(3)  There is correspondency between "a character recognition device 14" of Constituent Feature "b" of the accused device and "recognition means" of Constituent Feature B of the present invention.
(4) Hence, Constituent Feature "b" of the accused device conforms to Constituent Feature B of the present invention.

(III)  As to Constituent Feature C
(1) There is correspondency between "a series of texts (character signals) character-recognized by a character recognition device 14" of Constituent Feature "c" of the accused device and "unprocessed information recognized by said recognition means" of Constituent Feature C of the present invention.
(2) There is correspondency between "a text divided into sentences one by one" (a primary specifying step) or "a text delimited into sentences one by one (a preliminary specifying step) of Constituent Feature "c" of the accused device and "processed information corresponding thereto" of Constituent Feature C of the present invention.
(3) There is correspondency between "in accordance with operation of a computer program", "a function of inserting a return code (i.e., a function of dividing a text into sentences one by one) at a position following *a character which satisfies return code insertion conditions*" (primary specifying step) or "a function for delimiting (a series of texts [character signals] character-recognized by the character recognition device 14) on a sentence-by-sentence basis and calling them one by one" (preliminary specifying step) of Constituent Feature "c" of the accused device and "information processing means for processing" of Constituent Feature C of the present invention.
(4) Hence, Constituent Feature "c" of the accused device conforms to Constituent Feature C of the present invention.

(IV) As to Constituent Feature D
(1) There is correspondency between "present in a character-recognized text (character signal) outputted from a character recognition device 14" of Constituent Feature "d" of the accused device and "present in the recognized information outputted from said recognition means" of Constituent Feature D of the present invention.
(2) There is correspondency between "a character which satisfies return code insertion conditions" (primary specifying step) or "a character which satisfies one-sentence calling conditions (preliminary specifying step) of Constituent Feature "d" of the accused device and "terminator information" of Constituent Feature D of the present invention.
(3) There is correspondency between "in accordance with operation of a computer program", "a function of instructing insertion of a return code after *a character which satisfies return code insertion conditions*" (primary specifying step) or "a function of instructing a call of sentences after delimited one by one" (preliminary specifying step) of Constituent Feature "d" of the accused device and "processing execution instruction means for causing said information processing means to execute a processing operation" of Constituent Feature D of the present invention.
(4) Hence, Constituent Feature "d" of the accused device conforms to Constituent Feature D of the present invention.

(V) As to Constituent Feature E
(1) There is correspondency between "a machine translation system" of Constituent Feature "e" of the accused device and "an information processing apparatus" of Constituent Feature E of the present invention.
(2) Hence, Constituent Feature "e" of the accused device conforms to Constituent Feature E of the present invention.

2. Re:  Defendant's Pleading 1
(Re: restrictive construction in view of the description in the specification and whether or not the present is encompassed within the technical scope of the present invention)

(I) A term "means for" is used in claims of the present invention only because it is a common way of describing the invention in the technical field of the present invention.  Even if the present invention is described with "means for", the constituent features of the present invention are not functional claims.
(II) The constituent features of the present invention do not include any indefinite point for one skilled in the art in view of the state of art as of filing the present application.
(1) As to "input means" of Constituent Feature A, not only voice input means but also a keyboard, a mouse, a light pen, an image scanner (a part of OCR), etc. had already been well-known as of filing the application of the present invention.  Even if it would be construed that those are included in the "input means", it is within the scope obvious to one skilled in the art.
(2) As to "recognition means" of Constituent Feature B of the present invention, a character reading device and a voice recognition device or the like as recognition devices for recognizing a character (character information) and a voice (voice information) or the like had already been well-known as of filing the application of the present invention.  Even if it is construed that those are included in the "recognition means", it is within the scope obvious to one skilled in the art.
(3) As to "information processing means" of Constituent Feature C, terms 'information' and 'information processing' were used to represent the following concepts, respectively, as of filing the application of the present invention:  "information" refers to "a meaning which is given to data by a human based upon given rules", and "information processing" represents the concept of "acquiring information which meets a purpose from given information".  Not only data processing but also translation, graphics, characters and voice recognition or the like are included therein".  Therefore, there is no vague point in the concept of Constituent Feature C for one skilled in the art.
(4) As to "detection of terminator information" of Constituent Feature D, microcomputer application technology, computer program utilization technology, and further, technology for using terminator information for the purpose of control were developed in various fields as of filing the application of the present invention.  Therefore, even if it is construed that means for detecting terminator information to execute an operation includes a computer program utilization, it is within the scope obvious to one skilled in the art.

(III)   The Defendant argues that the each constituent feature of the present invention is comprehensive and abstract.  However, even if each of the constituent features may be a concept which includes extensive technical knowledge, it would not give rise to any problem if its meaning is clear from the description in the specification or the state of art as of filing the application of the present invention.

(IV) As set forth above, where each constituent feature of the present invention does not require a specifically restrictive construction, the accused device includes a structure as set forth in the Plaintiff*s Pleading 2 concerning Point at Issue 1 (List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Plaintiff).  Thus, all the constituent features of the accused device conform to those of the present invention.

3. Re:  Defendant's Pleading 2
(as to restrictive construction in view of the prosecution history and whether or not the accused device is encompassed within the technical scope of the present invention)

 The Defendant argues that the constituent features of the present invention should be construed in a restrictive manner on the basis of the description in the original specification of the parent application.  However, as has been stated at 2 (II) above, it would have been obvious for one skilled in the art to construe the meaning of each constituent feature as set forth in the above 2 (II) in the light of the parent application and the state of the art as of filing the application of the present invention.  Consequently, there is nothing that necessitates to accept the restrictive construction as argued by the Defendant.

4. Re:  Defendant's Pleading 3
(as to the restrictive construction based upon the known art and whether or not the accused device is encompassed within the technical scope of the present invention)

(I) The invention of Exhibit B1 is premised upon keyboard input.  However, there is no skilled person in the art who would recognize a keyboard as recognition means even if the keyboard is construed in a broad sense.  Therefore, the present invention is not considered as being entirely publicly known over the invention of Exhibit B1.

(II) In the inventions of Exhibits B8 and B9, "detection of punctuation marks" as argued by the Defendant is not performed.  Even if it is presumed that "detection of punctuation marks" is performed in the inventions of Exhibits B8 and B9, such detection is not utilized in subsequent operation (i.e., insertion of paragraphs).  Consequently, the present patent is considered as being entirely publicly known over the inventions of Exhibit B8 and B9.

(III)   The invention of Exhibit B17 is merely directed to improvement of a recognition rate in an optical character reading device which corresponds to recognition means of the present invention.  Further, the invention of Exhibit B17 does not include any structure corresponding to processing execution instruction means of the present invention.  Furthermore, in the invention of Exhibit B17, blanks for a given period of time (terminator information) is detected in information before recognition rather than after recognition.  Consequently, the present invention is different from the invention of Exhibit B17, and thus cannot be considered as being entirely publicly known.

(IV) From the foregoing, all the arguments concerning the restrictive construction on the basis of the publicly known art by the Defendant lack convincing grounds, and thus, there is not at all any necessity to restrictively construe the constituent features of the present invention.

5. Defendant's Pleading 4
(as to the restrictive construction from the standpoints of effect and advantageous results and whether or not the accused device is encompassed within the technical scope of the present invention)

(I) As to Removal of a Noise

(1) Discussion of the effect and advantageous results of the facts of the case 3 (I) also applies to a case in which information before recognition is directed to an image signal while information after recognition is directed to character code information.  That is, in an image scanner output, noises are prone to be produced as small black dots during copying or paper dirt.  However, such noise is not outputted through a character recognition device but is cut.  Therefore, only character code information as a result of the recognition is outputted by the character recognition device.  In such case, since terminator information targeted for detection becomes character code information, a detecting accuracy is substantially improved compared to the case of detection from information before recognition containing considerable noises.  Therefore, with the structure of the present invention, the same effect and advantageous results can be produced even in the case of a character, as in the case of a voice.
(2) The Defendant argues that in the case of character information, it is impossible to detect a period (a full stop), which is terminator information, in information before recognition.  However, for example, the following is possible.
A horizontal width of a character is detected, and a width of a period is subtracted from this horizontal width.  As a result of such subtraction operation, a portion where an output has ceased is detected as a period (a full stop).  The detection of the horizontal width of the character is, as disclosed in Exhibit A29, a generally well-known art that is employed when characters are cut out one by one from a character string in an OCR or the like.  Meantime, it is possible to detect a period (terminator information) existing in an image signal (information before recognition), but there are difficulties with such detection , and further, incorrect detection will increase because an image signal of a period does not reflect its exact shape due to paper "dirt" (paper noise), "fluctuations in sensitivity" of a light receiver element, "fluctuations in illumination" of a light emitting element (a noise of an observation system), "non-uniform concentration, a breakage, blur, and collapse" which will appear in a part of or the whole of a character printed out by a printer, "change or deformation" of character shape, "vertical displacement" or the like (a noise in a character generation process), or quantizing noise.
However, noises can be eliminated from even the above information containing noises if it is passed through the recognition device, and thereafter, converted into a character code signal indicative of a period.  In this regard, the detection can be achieved easily with a high accuracy because the target character signal to be detected has been processed to thereby become noiseless by the above-mentioned processing.

(II) Timely Instruction for Execution

 As has been described above, the present invention is featured in that terminator information after recognition is target information for detection, thereby producing effect and advantageous results as set forth in the facts forming a basis of the case 3 (II) hereinabove.
 In the present invention (Constituent Feature D), it is defined that terminator information is present "in information after recognition to be outputted from recognition means", and that there is recitation of giving an instruction to execute a processing operation "upon detection" of terminator information.  Therefore, detection and the instruction for execution are continuous.
 However, there is no necessity that "output from recognition means" and "detection of terminator information (and an instruction for execution)" are continuous.  The following case is also encompassed within the technical scope of the present invention:  that is, a case in which:
information recognized by the recognition means (including terminator information) is temporarily stored in a memory or the like and called from the memory at a proper time.  Then, the terminator information present in the called information is detected (here, the terminator information refers to "terminator information which is present in the recognized information outputted from said recognition means").
 Therefore, information outputted from the character recognition device of the accused device is temporarily stored in a RAM and then, is called.  This information is nevertheless the same as the information after recognition outputted from the character recognition device, and thus corresponds to "the recognized information outputted from the recognition means" of Constituent Feature D of the present invention.

[Defendant's Pleading]

1. Restrictive construction on the basis of the description of the specification and whether or not the accused device is encompassed within the technical scope of the present invention

(I) All of the constituent features of the present invention are expressed as so-called functional claims, using terms such as "functions to be achieved" and "means".  Thus, each of constituent features is expressed in an extremely ambiguous and comprehensive manner.  In such case, the technical scope should be limited to the extent such that the invention is easily carried out by or obvious to one skilled in the art in view of the description under the Detailed Description of the Invention in the specification which was disclosed for the purpose of disclosing the invention.

(1) In the present specification, only voice input is disclosed as means for "inputting information" of Constituent Feature A.  Therefore, Constituent Feature A should be limited to a voice inputting device.
(2) In the present specification, only voice recognition is disclosed as means for "recognizing information" of Constituent Feature B.  Therefore, Constituent Feature B should be limited to the voice recognition device.
(3) In the present specification, only an electronic translation device is disclosed as "information processing means" of Constituent Feature C.  Therefore, Constituent Feature C should be limited to the translation device.
(4) In the present specification, there is disclosure as to only detecting that a voice input blank exceeds a predetermined time by employing a time constant circuit, as "upon detection of terminator information" of Constituent Feature D.  Therefore, Constituent Feature D should be limited to causing the electronic translation device to make electronic operation every time it is detected that the voice input blank exceeds a predetermined time by employing the time constant circuit.

(II) In the Plaintiff*s Pleading concerning the significance of each constituent feature, the Plaintiff has the benefit of the constituent features of the present invention which are abstract and comprehensive, and thereby attempts to unduly broaden the content of its meaning beyond the extent that is understandable by a person skilled in the art in order to carry out the invention on the basis of the description in the specification.

(III)  The accused device generally has a structure as identified in the Plaintiff*s Pleading 2 concerning Point at Issue 1 (List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Defendant).  Therefore, the constituent features of the accused device do not conform to any of those of the present invention.

2. Restrictive construction in view of the prosecution history and as to whether or not the accused device is encompassed within the technical scope of the present invention

(I) In the specification originally attached to the parent application of the present invention, only the following contents were disclosed.  Thus, as long as the application of the present invention is a divisional application from the parent application, the meaning of each constituent feature should be construed as being limited to:

(1) The "information inputting means" of the present invention corresponds to only a voice input device.
(2) The "information recognition means" of the present invention corresponds to only voice recognition using a voice analyzing device.
(3) The "information processing means" of the present invention corresponds to only an electronic translation device.
(4) "Detection of terminator information" of the present invention corresponds to only detecting whether the blank time is long or short using a time constant circuit, while the voice input terminator is identified with the fact that the blank time in voice input is longer than a predetermined time.

(II) If the significance of each constituent feature is broadly construed as argued by the Plaintiff, it naturally departs from the scope disclosed in the specification originally attached to the parent application.  As a result, the patent application for the present invention was the divisional application illegally filed.

(III)   The accused device generally has a structure as identified in the Defendant*s Pleading 2 concerning Point at Issue 1 (List of Correspondency between Constituent features of the Accused Device and Those of the Present Invention submitted by the Defendant, attached hereto).  Thus, the constituent features of the accused device do not conform to any of the constituent features of the present invention.

3. Restrictive construction on the basis of the publicly known art and as to whether or not the accused device is encompassed within the technical scope of the present invention

(I) The following techniques were publicly known as of filing the patent application of the present invention:

(1) The art relating to a Japanese word processor as described below:
in inputting from a key, a character corresponding to the Kana-character key is identified, the character code signal is generated and a Japanese language sentence expressed in Kana-characters is stored; and then
when a punctuation mark is inputted, a portion corresponding to Kanji characters in the Japanese sentence inputted in Kana-characters until then is converted into Kanji characters (Exhibit B1).

(2) An art relating to a device for mechanically analyzing the construction of Japanese sentences, wherein commas are detected in Japanese sentences; a paragraph level code is assigned by a paragraph signal inserting circuit of a paragraphing circuit; and thereafter, the paragraphing result is written into an analysis result storage area by a writing circuit (Exhibit B8).

(3) An art relating to an apparatus for mechanically analyzing construction of the Japanese sentences and, according to a result thereof, arranging and displaying the sentences in a form easy to read, wherein a paragraph level code is assigned to the paragraph at the last word of each paragraph by means of the paragraphing circuit, a paragraph level code from an analysis result storage area is detected, and a required control code (return code) is inserted into a document character code string by an analysis result output circuit (Exhibit B9).

(4) An art relating to an optical character-reading apparatus which optically reads information regarding characters or signs and the like by moving the apparatus over a paper including the characters or sings and the like while gripping it by hand, wherein, if a character identification output is not present for a given period of time or more, an operation for adding LF signal indicative of completion of reading one line is performed (Exhibit B17).

(II) Where each constituent feature of the present invention is construed broadly as done by the Plaintiff, the aforementioned known techniques shall be also encompassed in the scope of the present invention.  Thus, the present invention is considered as entirely known art.  In defining the technical scope of the present invention, it should be construed that the features disclosed in these publicly known techniques are not included in the present invention even if such techniques are literally included in the claim of the present invention.  Consequently, the significance of each of the constituent features of the present invention should be construed as limited to the significance which has been set forth in the above item 1.

(III)   The accused device generally has a structure as identified in the Defendant's Pleading 2 concerning Point at Issue 1 (List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Defendant, attached hereto).  Therefore, the constituent features of the accused device do not conform to any one of the constituent features of the present invention.

4. Restrictive construction in view of the effect and advantageous results and as to whether or not the accused device is encompassed within the technical scope of the present invention

(I) As to Noise Elimination
 In the case of character information on language sentence, unlike from the voice information, the effect and advantageous results as set forth in the facts forming a basis of the case 3 (I) are not produced.

(1) In character information, unlike the voice information, the presence of a peripheral noise per se is extremely abnormal.  Therefore, the character information is essentially different from the voice information in which the presence of such noise is normal.
 Consideration is given to the following case in which a peripheral noise which rarely generates is present.
In extracting a period included in a character string in a state in which a bit of peripheral noise before recognition is included, a degree of extraction (discrimination) of the period in the characters is hardly affected by the presence of such a slight background noise.  In other words, unlike a case in which a blank after voice recognition (no-signal state) is detected, there is no technical problem, or there hardly exists a difference between before and after recognition, so that the effect and advantageous results of the present invention are hardly achieved.

(2) Voice information is featured in that a long or short blank as terminator information remains as is, irrespective of before or after voice recognition and terminator information is detectable irrespective of before or after voice recognition (thereby making it possible to make comparison between a detecting accuracy before and that after the recognition).  On the contrary, in the case of the character information in the accused device, assuming that a period is terminator information, it is through character recognition that a period can be identified distinctly from other characters, that is to say, a period before character recognition is a mere image signal, like other characters.  Before the recognition, there does not exist any art for identifying a character and a period, irrespective of the presence or absence of a noise.  In short, the feature resides in that a character and a period cannot be recognized before the character recognition even if a noise is absent.  As has been discussed above, "a period" per se as being "detectable" "terminator information" is not present naturally before recognition.  Thus, it is impossible to recognize a relationship such that "recognition" is performed for the purpose of comparison with information before the recognition to eliminate the noise which becomes an obstacle in detecting a period, thereby enhancing a period detecting accuracy.

(3) The effect and advantageous result unique to the present invention reside in that "the detecting accuracy can be improved substantially by making comparison with a case wherein terminator information is detected from information before recognition containing a large amount of noises", and thus "a noise" in question should be a noise causing an obstacle in detecting "a blank" as "terminator information".  Moreover, "a noise" in question does not mean such noise that causes "incorrect recognition", but solely means "a noise which does not cause an obstacle to recognition" and "a noise within a noise allowance for recognition".
 However, the accused device is featured in that both characters and periods are converted into code signals by "comparison of analogy" and then, such code signals of periods are discriminated from other code signals of characters in character-recognizing the image information.  Thus, as far as the noise amount is within the limit that permits to recognize the inputted information (i.e. within the aforementioned "the noise allowance for recognition"), such noise would not hinder discrimination of periods (not limited thereto).  In short, the accused device does not include "a noise" so called in the present invention.  As studied earlier, the "noise" so called in the present invention requires:
"a noise not to become a hindrance to recognition (incorrect recognition)"; as well as
 a noise to become a hindrance to detection of terminator information (a blank)."
 In view of the foregoing, the "noise" so called in the present invention does not become a hindrance to discrimination of periods in the accused device, so that any of the technical problems per se is not present or the effect and advantageous results unique to the present invention are not produced.

(II) As to Timely Instruction for Execution

(1)  From the viewpoints of the effect and advantageous results of the present invention as admitted in the facts forming a basis of the case 3 (II), Constituent Feature D is structured such that operation of recognition means (with a time lag) and operation of information processing means are performed continuously and sequentially in time relating to an information portion having a blank at its terminator (that is, one sentence).  Therefore, it should be construed that an output after recognition from the recognition means is limited to those supplied to the information processing means directly and continuously in time.
(2) Contrary to the above, the accused device is not so structured that an output from recognition means having terminator information at its terminal is supplied to the information processing means directly and continuously in time, nor structured such that information processing of information processing means is executed with a timing such that the terminator information is detected from the output of the recognition means.  In the light of the above, the accused device does not produce an advantageous result in that these two operations (recognition and information processing) can proceed sequentially without a timing conflict.
 In terms of "insertion of the return code" in registering operation where the Plaintiff dominantly argues as an infringement, all of the character code data outputted from a character recognition device is temporarily stored and accumulated in an RAM in a main body, and after completion of all outputs, registering operation will be executed responsive to the operator*s instructions to select and execute registering operation.  In short, in the accused device, the output of all the information from the character recognition device is completed and stored in the RAM.  Thereafter, the information is targeted for search, and it is judged as to whether or not a predetermined return code insertion position is contained in the information.  Thus, the accused device is not directed to a device wherein terminator information is detected with a timing at which the recognition result is outputted from the character recognition device.
 Further, the aforementioned discussion can also apply to a case wherein a function for "calling sentences distinctively one by one" in a Japanese sentence calling operation, which is allegedly a preliminary specifying operation according to the Plaintiff, is grasped as information processing means.  That is, the Japanese sentence calling operation is performed according to the operator*s instruction, subsequent to an operation that a character-recognized text is registered in an original file by the operator*s instruction for selecting and executing the registering operation.  Thus, it is not structured such that that an output from the recognition means having terminator information at its terminal is supplied to the information processing means directly and continuously in time, nor such that information processing of the information processing means is executed with a timing at which terminator information is detected from the output of the recognition means.  Yet further, it is not structured so as to produce an advantageous effect that two operations (recognition and information processing) can proceed sequentially and timely.
 Meantime, the accused device retrieves character code data from various sources, such as a file created by a word processor other then the accused device and a file created by keyboard input and the like, besides a file registering outputs from character recognition device as explained above, and then, calls sentences one by one under the single-sentence calling condition.
Thus, the accused device is different in that "in detecting terminator information present in information after recognition to be outputted from recognition means, processing operation is executed".
(3) Thus, the constituent features of the accused device do not conform to Constituent Feature D of the present invention.

III. Point at Issue 3
(as to the amount of damages payable to the Plaintiff in the present case where the Defendant is obliged to indemnify the Plaintiff for the damages)

[Plaintiff's Pleading]

1. The Defendant sold about 1,800 units of the accused devices at a unit price of about 7,000,000 Yen during a period from February 21, 1994 when the patent right of the present invention was assigned to the Plaintiff, up to December 20, 1995, and obtained profits therefrom amounting to at least 367,000,000 Yen.

2. Hence, while the Plaintiff possesses a right to claim for the same damages mentioned above against the Defendant, the Plaintiff actually claims for payment of 75,000,000 Yen as the partial damages against the Defendant.

[Defendant's Pleading]
 The Defendant contests the Plaintiff's Pleading.

IV. Court Decision as to Points at Issue
 There is a dispute between the Plaintiff and the Defendant as to how the structure of the accused device should be specified (Point at Issue 1).  Here, study will be made as to the restrictive construction in view of the effect and advantageous results and as to whether or not the accused device is encompassed within the technical field of the present invention from among the issues in Point at Issue 2 (Defendant*s Pleading 4) based upon the Plaintiff*s Pleading concerning Point 1 at Issue.

I. As to the Features of the Present Invention

1. According to the above facts forming a basis of the case, Exhibits listed later and the entire gist of the pleadings, the following facts are admitted.

(I) The patent application for the present invention is a divisional application from the parent application (Japanese Patent Application No. Sho 55-176296).

(II) The content of the invention in the parent application (after amendment) is as follows (Exhibit B2):

(1) Scope of Claim for a Patent (Claim 1)
"An electronic translation apparatus comprising:
 receiving means for receiving information on a source language to be translated, the information being sent from afar;
 translating means for translating, from the information on the source language to be translated which was received by said receiving means, into information on a translation language corresponding thereto;
 output means for outputting the information on the translated language which was translated by said translating means;
 termination detecting means for detecting a termination of the information on the source language to be translated which was received by said receiving means; and
 translation instructing means for, when the termination is detected by said termination detecting means, causing said translating means to execute translating operation."  (Amended Claim in an Amendment dated on May 31, 1987)

(2) Problems to be Solved by the Invention and Advantageous Effect of the Invention

a. "The present invention relates to an improvement of an electronic translation apparatus.  In particular, the present invention relates to an electronic translation apparatus for detecting a termination of a source language input signal, thereby automatically starting electronic translation."

b. In a conventional electronic translation apparatus:
"A speaker had to manually operate a translation switch to obtain a translated language every time a source language is inputted.";
"In an electronic translation apparatus capable of voice input equipped with a voice recognition circuit, a source language can be directly voice-inputted, thus dispensing with manual operation.  However, the manual operation is indispensable for instructing a translation operation."; whereas
"an object of the present invention is to eliminate such disadvantages."

c. "The present invention is equipped with receiving means for receiving source language information to be translated sent from afar; translating means; termination detecting means for detecting a termination of the source language information; and translation instructing means for causing to execute a translating operation upon detection of the termination.  Thus, the present invention has significantly large practical value in the following respects:
automatic translation reception can be performed at a site distant from a transmitter to which source language information is inputted without requiring manual operation;
when voice recognition means is employed, automatic translation broadcasting and automatic translation telephone or the like can be achieved; and
a translation apparatus and a translation communications apparatus can be automated.

(III)  The Official Notice of Reason for Rejection was mailed on October 6, 1987 (Exhibit A21-1), and thereafter, Final Decision of Rejection was notified on October 27, 1988 (Exhibit A21-2) in connection with the parent application.  The reason for the Final Decision for Rejection was as follows:
"..., execution by voice termination is publicly known (over Reference 2), and in general, it is an obvious matter within the scope of a common practice to execute an operation by detecting information such as the information on data file termination (EOF), data transmission termination (EOT) and the like."
 This Final Decision for Rejection became conclusive as it was.

2. In view of the content of the parent application, the reason for Final Decision for Rejection, and further, comparison between the invention of the parent application and the present invention, it shall be appropriate to construe that the present invention was patented focusing upon the following two advantageous effects, i.e., (i) noise elimination and (ii) timely instruction for execution, as stated in the facts forming a basis of the case hereinabove, and these advantageous effects can be produced in the following manner, that is:
Target information for detecting terminator information is information after recognition of inputted information, but not information before recognition of the inputted information on the premise that the terminator information contained in the inputted information is detected to automatically execute the information processing operation that is well known to the public.
It shall be further appropriate to construe, where the essential constituent features of the present invention is as set forth above, that the art which does not produce such effect and advantageous results is not encompassed within the technical field of the present invention even if it appears that such art literally conforms to the constituent features of the present invention.

II. As to the Advantageous Result of Noise Elimination

 The noise elimination advantageous result of the present invention is as has been described above.  That is, the terminator information detecting accuracy in detecting terminator information from information after recognition, is improved in comparison with the detecting accuracy in detecting terminator information from information before recognition.  This is because noises which prevent accurate detection of terminator information are eliminated in the recognition process.  The present invention is featured in that the detecting accuracy is enhanced in comparison with as the premise, the case where terminator information is detected from the information before recognition.  To say that the accused device produces such effect and advantageous result, the following requirement must be met.  That is, it is required to make it possible to detect terminator information from information before recognition, even in the case where characters are inputted as image information, and punctuation marks or the like are detected as in the accused device in the Defendant*s Pleading.
 In the meantime, according to the Plaintiff's Pleading (List of Correspondency between Constituent feature of the Accused Device and Those of the Present Invention submitted by the Plaintiff, attached hereto), in the accused device, information before recognition designates an image signal (image information) obtained by reading characters by means of an image scanner; information after recognition designates a character signal (character code information) recognized by calculating the above image information based on a degree of analogue; and terminator information designates "characters satisfying the return code insertion condition" or "characters satisfying the single-sentence calling condition" (hereinafter, referred to as "punctuation marks").  The Plaintiff argues that it is possible to detect punctuation marks or the like from image information before recognition by showing Exhibit A29.
 Though the specification of the present case has been studied throughout, there was no specific reference to a case where characters are inputted as image information.
 In addition, Exhibit A29 is an excerpt of "MOJI NINSHIKI GAIRON (General Study of Character Recognition)" compiled by Shinichiro HASHIMOTO, published on March 20, 1982.  At a passage from page 62 therein, there is explanation on "a process for detecting a position of a circumscriptive frame of a target character pattern, that is, right and left ends of the pattern, and then, coordinates of top and bottom ends" as means for "cutting out characters".  Specifically, there is primarily explained a method for detecting the right and left ends of the pattern in which these right and left ends are obtained from a projected character pattern on the X-axis, illustrating each of the cases when the projection is continuous or discontinuous, and when adjacent characters are superimposed.  The Plaintiff argues, based upon such disclosure, that there is "a method comprising detecting a horizontal width of a character, subtracting a width of a period from the horizontal width, and then, as a result of the subtraction, detecting a portion devoid of output as a period", thereby making it possible to detect a period that is terminator information even in information before recognition.  However, the above-described content relates to mere art of extracting a region for one character by detecting horizontal and vertical widths as preprocessing for character recognition.  Therefore, art for detecting the punctuation marks could not be directly derived from such method for detecting a horizontal width of a character, nor one skilled would find it obvious to reach such art.
 Furthermore, as of December in 1992 which is assumed as a date when the Defendant started manufacturing and selling the accused device, there was no good evidence which is sufficient enough to substantiate admitting that the art of detecting punctuation marks or the like immediately from character information inputted as image information.
 In the light of the foregoing, as in the accused device, in the case where characters are inputted as image information to thereby detect the punctuation marks, it is impossible to arrive at an idea of detecting terminator information from information before recognition.  Thus, it is construed, even if terminator information is detected from information after recognition, that there would be no room for producing effect and advantageous results of improved detecting accuracy.
 Consequently, even on the premise what the Plaintiff argues concerning the accused device, the accused device is not encompassed with the technical scope of the present invention.

III. As to the Timely Instruction for Execution

1. An advantageous effect of timely instruction for execution according to the present invention is as described below:
In comparison with a case in which terminator information is detected from information before recognition, when terminator information is detected from information after recognition, recognition operation by recognition means and input operation into information processing means have been already completed upon detecting the terminator information.  Thus, when an automatic processing execution instruction is supplied to the information processing means, it is not necessary to consider a time lag required for recognition by recognition means.
 In the meantime, when terminator information is detected from information before recognition, the following issue is raised:
There is need to consider time lag in instructing the execution to the information processing means.  This is the issue which is inherent to a situation in which recognition and information processing are performed sequentially, continuously in time for the reason stated below.  When recognition and information processing are discontinuous in time as a case in which information processing is performed by making a batch of information after recognized instead of performing the recognition and the information processing sequentially, continuously in time, information processing is not performed until recognition of all inputted information has been completed even when terminator information is detected from information before recognition.  Thus, it is impossible or meaningless to execute information processing with the above detection timing.  In order to detect terminator information to automatically execute information processing, it inevitably follows that there is no alternative but to provide a structure such that terminator information is detected from information after recognition.  Therefore, where execution of automatic processing is instructed to the information processing means under the situation where recognition and information processing are not performed sequentially, continuously in time, it should be appropriate to construe that the advantageous effect of the present invention cannot be produced under such situation because this situation does not permit to make a comparison between a case using a timing with which terminator information is detected from the information before recognition and a case using a timing with which terminator information is detected from the information after recognition.
 In this regard, the Plaintiff argues that detection of terminator information and an instruction for executing the information processing need to be continuous, whereas recognition, and detection of terminator information and an instruction for executing the information processing do not need to be continuous.  In fact, based upon the mere literal meaning of the description that terminator information is detected from information after recognition, it may be possible to construe that a process from recognition to information processing does not need to be continuous.  However, as has been described above, the present patent was patented, focusing upon the fact that the present invention resolved the problem associated with the case in which terminator information is detected from the information before recognition.  Hence, it should be construed that the present invention is premised upon the requirement or the prerequisite that the situation exists which possibly gives rise to a problem associated with the case of detecting terminator information from the information before recognition.  In the case of a situation in which recognition and information processing are not sequential, continuous in time, the above-mentioned prerequisite is not met, as has been shown above.  Therefore, the above Plaintiff's argument is unacceptable.

2. Then, in order to determine whether or not the accused device meets the prerequisite, study will be made.  According to the Plaintiff's Pleading (List of Correspondency between Constituent Features of the Accused Device and Those of the Present Invention submitted by the Plaintiff attached hereto), the accused device operates as follows:  (i) after character image information is inputted by an image scanner; (ii) the information is recognized by the character recognition device, and it is converted to a character signal (character information) as a result of the recognition; and (iii) when an operator selects interactive processing and selects a specific document file, then,
the accused device subsequently operates as follows:  (a) upon detecting that "a character which satisfies the return code insertion condition is present in a series of texts (character signals) which are character-recognized by the character recognition means", an operation for inserting return code after the detected character is executed, or (b) upon detecting that "a character which satisfies the single-sentence calling condition is present in a series of texts (character signals) character-recognized by the character recognition means", an operation for instructing to call sentences distinctively one by one is executed.  As is evident from the Plaintiff's argument stated above, in the accused device, after a series of texts targeted for recognition are entirely character-recognized, periods or the like are detected to insert return codes.  This means that recognition and information processing are not always performed sequentially and continuously in time.  This point is more clearly explained in the Plaintiff*s Pleading (Plaintiff*s Pleading 4 (II) concerning Point at Issue 2).

3. Therefore, even premising the Plaintiff's argument concerning the accused device, recognition and information processing are not performed sequentially and continuously in time in the accused device.  Thus, the accused device does not conform to Constituent Feature D of the present invention.

V. Conclusion
 As a result of the foregoing study, there is no reasoning in the Plaintiff*s demand even if the remaining issues are not studied so far.  Thus, the decision is the same as stated in the main text in the present case.

(The hearings were concluded on December 8, 1998.)
Civil Department 21, Osaka District Court
Kazuo KOMATSU, Presiding Judge
Hiroyuki TAKAMATSU, Judge
Keiko SETO, Judge