US 4682243

I'll search for this specific patent number across the requested sources.

The patent page is confirmed. Let me check the USPTO database and CAFC dockets specifically.

US Patent 4,682,243 — Summary

Note on sources and uncertainty: My searches reached Google Patents (authoritative full text) and Justia, plus general web results. I do not have authenticated direct access to USPTO PatentCenter or to the CAFC 2026 docket system from these tools, so the litigation check below is based on public web indexing rather than a direct docket query. Where I could not verify something, I say so.

Bibliographic data

Abstract (as issued)

In reproducing images in a desired magnification ratio by storing in a memory image data obtained from an original and reading out the data from the memory, moire patterns are reduced by a sampling frequency constant in accordance with a magnification ratio that yields good reproduction quality. Image data synchronized with the sampling frequency are averaged among a number of pixels corresponding to the desired magnification ratio, for storage.

Plain-language overview of the independent claims (claims 1, 2, 3, 6)

Claim 1 — Method (the broadest independent claim). A method for controlling magnification ratio in an electronic image reproduction system, with four steps:

• (a) scan an original image;
• (b) sample it in synchronism with a sampling pulse signal of a constant frequency f_s that is independent of the selected magnification ratio, producing digital image data;
• (c) average the digitized data over a number of pixels corresponding to the magnification ratio; and
• (d) write the averaged data to a memory synchronized with a write pulse signal of frequency f_in = K₂·M·f_s, where K₂ is a constant and M is the magnification ratio.

Plainly: keep the sampling clock fixed and instead reduce data volume by averaging groups of pixels, so the write clock tracks the magnification ratio.

Claim 2 — Method (shift-register / power-of-two variant). Another method with these steps:

• (a) scan; (b) sample in synchronism with a sampling pulse of frequency f_sp to get digital data for writing to memory;
• (c) generate a write pulse of frequency f_in = f_sp / 2ʳ (r an integer);
• (d) average the data every 2ʳ pixels by dividing (shifting) a sum of 2ʳ pixel values;
• (e) write the averaged data to memory in synchronism with the write pulse at frequency f_s0 × K₂·M, where M is the magnification ratio and f_s0 is a constant sampling frequency independent of magnification; and
• (f) read the memory with a constant read pulse signal (independent of magnification) for the recording process.

Plainly: the divider is replaced by a simple bit-shift, which only works cleanly for averaging in powers of two (2ʳ pixels).

Claim 3 — System (first apparatus claim). A variable-magnification image reproduction system comprising:

• (a) means for scanning an input image;
• (b) sampling means synchronized to a constant-frequency f_s sampling pulse independent of magnification;
• (c) A/D converter means for digitizing;
• (d) averaging means for averaging the digitized data over a number of pixels corresponding to the magnification ratio;
• (e) means for generating a write pulse whose frequency is related to the sampling frequency and the magnification ratio;
• (f) means for generating a constant-frequency read pulse independent of magnification;
• (g) memory means written synchronously with the write pulse and read synchronously with the read pulse; and
• (h) means for applying read and write pulses to the memory.

Claim 6 — System (power-of-two / shift-register variant). A variable-magnification electronic image processing system comprising:

• (a) means for scanning an original;
• (b) sampling means producing corresponding digital image data;
• (c) averaging means that averages over 2ʳ pixels by dividing (shifting) the sum of image data for 2ʳ pixels;
• (d) write-pulse generating means producing a write pulse of frequency f_sp = f_s0 × K₂·M (K₂ constant, M magnification ratio, f_s0 constant reference sampling frequency independent of magnification);
• (e) memory means written synchronously with the write pulse and read synchronously with a constant read pulse; and
• (f) sampling-pulse generating means producing the sampling pulse at frequency 2ʳ·f_sp.

(Dependent claims: 4 and 5 depend from claim 3; 7, 8, 9 depend from claim 6. Claim 4 details the divider-type averaging circuit — adder, register, data selector, preset counter, frequency divider; claim 7 details the shift-register/adder/register/data-selector averaging circuit; claims 5 and 8 recite a pulse signal distributor; claim 9 recites a pulse signal distributor plus a ÷1/2ʳ frequency divider.)

Technical gist

The invention addresses moire that appears when the memory write/sampling frequency is dropped well below the read frequency for reduction. Rather than lowering the sampling frequency (which required changing the scanning beam diameter and aperture size), it holds sampling frequency high and constant, averages pixel data over a number of pixels set by the magnification ratio, and varies the write frequency so the ratio f_s/f_out stays constant. Two embodiments are shown: one using a divider (FIG. 3) and one using a shift register plus frequency divider (FIG. 5).

CAFC 2026 docket check

• A targeted search for "4682243" / "4,682,243" in Federal Circuit / appeal contexts returned no results tying this patent to any Federal Circuit litigation, and no CAFC 2026 docket entry for it.
• Consistent with that, the patent's legal-status record shows it lapsed in 1999 (failure to pay maintenance fees) and was recorded as expired — making current appellate litigation on this patent improbable.
• Caveat: I could not run a direct, authoritative query against the CAFC docket (CM/ECF) or USPTO PatentCenter with these tools. This is a "no indexed hits found" result, not a certified absence. If you need court-grade confirmation, query the CAFC PACER/CM-ECF docket and USPTO PatentCenter directly for application 06/716,686.

No auto-correction was applied: the number is treated literally as 4,682,243 (US4682243A), and results for other numbers were excluded.

I'll search the litigation databases and the patent number for any litigation records.

The broad searches returned mostly unrelated " '243 " patents. Let me run more targeted queries against litigation-specific sources.

Litigation Search Report — U.S. Patent No. 4,682,243

Patent identified: U.S. Patent No. 4,682,243 ("Magnification ratio conversion in image reproduction"), inventor Fumihiro Hatayama, assignee Dainippon Screen Mfg. Co., Ltd., granted July 21, 1987, expired July 21, 1999 due to non-payment of maintenance fees.

Bottom line

No litigation involving U.S. Patent No. 4,682,243 was found. None of the searches returned any case in which this patent is identified as an asserted patent, an invalidated patent, a counterclaimed patent, or the subject of any declaratory judgment, IPR, or appeal.

What I searched

I ran targeted queries against general web indexes that surface federal court dockets and litigation aggregators (CourtListener, Justia, PlainSite, Casetext, Stanford NPE database, PTAB-related documents), using the specific identifiers "4,682,243," "4682243," plus the assignee name "Dainippon Screen." Every query was constrained to the exact number; I did not accept results for similar numbers.

Critical disambiguation (do not conflate these)

The searches surfaced several cases and several different patents that are colloquially called "the '243 patent" but are NOT U.S. 4,682,243. I am flagging these explicitly so they are not misattributed:

None of the above involves U.S. 4,682,243.

Cases involving the assignee (Dainippon Screen), but not this patent

The searches did return litigation involving Dainippon Screen Mfg. Co. as a party — but all concern unrelated technology (semiconductor wet-processing/cleaning/drying and prepress imaging hardware), not the magnification-ratio/moire patent at issue here:

• Dainippon Screen Mfg. Co. v. CFMT, Inc., 142 F.3d 1266 (Fed. Cir. 1998) — declaratory judgment action; appellate decision on personal jurisdiction.
• Mattson Technology / CFM Technologies v. Dainippon Screen (N.D. Cal.) — wafer-drying patent suit; jury verdict for Mattson, settled 2002 for $40M.
• FAS Technologies, Ltd. v. Dainippon Screen Mfg. (N.D. Cal., 3:00-cv-01879) — patent/misappropriation suit; judgment for Dainippon Screen including fees.
• Creo Products Inc. v. Dainippon Screen Mfg. Co., 535 U.S. 1109 (2002) — cert granted, vacated and remanded in light of Festo.
• Raster Printing Innovations, LLC v. Dainippon Screen Graphics (USA), LLC (N.D. Ill., 1:15-cv-03991) — NPE suit.

I found no indication that U.S. 4,682,243 was ever asserted or challenged in any of these or any other proceeding.

Why this result is plausible

The record shows U.S. 4,682,243 lapsed in 1999 for failure to pay maintenance fees and was recorded as expired (37 CFR 1.362). Its term ended before the modern NPE assertion wave, and it belongs to a niche prepress image-reproduction field. Both factors make litigation highly improbable.

Caveats and limitations (read before relying on this)

1. This is a "no indexed hits found" result, not a certified absence. My tools query public web indexes, not a court-grade system. I could not run an authenticated, direct query against PACER/CM-ECF, the CAFC docket, USPTO PatentCenter, or Docket Navigator / Unified Patents' litigation database directly.
2. Coverage limits. District court complaints from the 1980s–1990s (when this patent was live) are frequently not digitized or indexed, so an old, resolved infringement action — if one existed — might not surface in these searches.
3. Exact-number discipline was maintained. No auto-correction was applied; the identifier is treated literally as 4,682,243 (US4682243A). Results for other " '243 " patents were excluded from the litigation findings.
4. For court-grade confirmation, the following direct sources should be queried with application number 06/716,686 or patent number 4,682,243:
   • USPTO PatentCenter / Patent Examination Data (assignment and status history);
   • PACER / CourtListener RECAP for district court dockets;
   • CAFC CM-ECF docket search;
   • Unified Patents litigation caselist (https://portal.unifiedpatents.com/litigation/caselist).

Conclusion: Based on the searches available to me, no known litigation involves U.S. Patent No. 4,682,243, and the cases that did appear were either about different " '243 " patents or about Dainippon Screen as a party in unrelated technology disputes.

I'll verify the canonical (empty) PTAB list against public sources before writing this up.

Proceedings overview

Total AIA trial proceedings on file for US 4,682,243: ZERO. Breakdown: 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution denials — because no petition has ever been filed. The structured "PTAB proceedings on file" block (USPTO Open Data Portal ingest) returns an empty set, and independent web searching surfaced no IPR, PGR, or CBM bearing on this patent. Bottom-line defensive posture for a defendant: there is no PTAB record to lean on — but there is also no live patent to defend against. US 4,682,243 lapsed for non-payment of maintenance fees effective 1999-07-21 and was recorded as expired (later confirmed by the 2018-01-31 STCH discontinuation event). All nine claims stand as issued and untested, but unenforceable-by-expiration. The correct defense posture is a standing/expiry objection, not an invalidity one.

No proceedings to report

The per-proceeding template does not apply, because there are no proceedings. I am stating this explicitly rather than leaving a blank, since "zero IPRs" and "I couldn't find them" are different claims.

What I verified (2026-10-01):

• Canonical source: The "PTAB proceedings on file" structured block in this prompt — drawn from the USPTO ODP API — contains no entries for US 4,682,243 / application 06/716,686.
• Corroborating search: Queries for the patent number in PTAB contexts returned nothing on point. Every hit using the string "243 patent" resolved to a different patent (see the false-positive warning below).
• Consistency check against the prior section: The earlier-generated CAFC check likewise found no Federal Circuit litigation. The two results are mutually consistent — no PTAB trial means no appeal from an FWD.

What I could not verify with these tools (stated so the record is honest, not hedged):

• I do not have authenticated query access to PTAB E2E / PTABX or to a direct ODP call in this session; the empty list rests on the provided structured block plus negative search results.
• Pre-AIA contested proceedings are outside the ODP AIA-trial dataset. US 4,682,243 has a 1984 priority date and issued in 1987, so it was never eligible for CBM (CBM was limited to financial-services patents) and inter partes reexamination (available only for patents filed on or after 1999-11-29) is also inapplicable. Ex parte reexamination would have been legally available, and the ODP "AIA trial proceedings" block would not necessarily list it. I found no evidence of any reexamination in the file history or the public record, but I cannot rule one out from a contemporaneous official filing receipt. If you need court-grade certainty, pull the full file wrapper for 06/716,686 via USPTO PatentCenter and search PTAB E2E by patent number.

False-positive hazard — do not conflate these two patents

This is the single most useful thing in this section. US 4,682,243 (Hatayama / Dainippon Screen, image magnification) is frequently confused in search results with US 7,682,243 (NHN Corporation / Game and Technology Co. ("GAT"), video-game data), which has real PTAB activity — including a petition by Activision Blizzard, Inc. and Blizzard Entertainment, Inc. challenging claims 1–8 of the '243 patent on § 103(a) grounds over JP2000-135372 ("Matsui") and other art (IPR petition, PTABX).

Both are informally "the '243 patent." They are unrelated in subject matter, owner, and era. If a demand letter or an internal docket flags "the '243 patent," confirm the number against the front page before importing any PTAB history. Attributing the GAT/Activision record to Dainippon Screen's patent would be a material error in either direction — it would either manufacture a false invalidity story or falsely signal that this patent is being actively litigated.

Strategic summary

Claim-by-claim status — untested (not canceled, not sustained). No claim of US 4,682,243 has ever been adjudicated by the Board. Independent claim 1 (the broad method claim with the constant-f_s sampling step and the f_in = K₂·M·f_s write-frequency relation) and independent claim 2; independent apparatus claims 3 and 6; and dependent claims 4–5 and 7–9 all stand exactly as granted. There is no narrowing IPR certificate, no adverse judgment, and no disclaimer. That sounds like a strong patent; it isn't, because of expiry — and because "never challenged" here reflects disinterest, not hardened validity. This is a 1987 patent in an obsolete drum-scanner architecture that stopped being asserted over a quarter-century ago.

Estoppel landscape — § 315(e)(2) is a non-issue. Because no IPR, PGR, or CBM was ever instituted, no petitioner and no privy is subject to IPR estoppel. There is no § 315(e)(2) bar, no § 325(e)(2) bar, and no estoppel chain to trace. Practically, this means a defendant would face zero estoppel friction in the prior art — the entire field is open. The corollary is that there is also no benefit to harvest: no petitioner has already spent the money building an invalidity record you could borrow, and no FWD gives you a pre-litigated roadmap. Everything would have to be built from scratch, which is exactly why nobody has bothered.

Pattern signals — absence is the signal. There is no repeat petitioner (no analogue to the serial-filing campaigns you see on heavily asserted patents), no defensive aggregator such as Unified Patents in the chain, and no Patent Owner appeals because there were no adverse decisions to appeal. USPTO assignment records show the patent remains with the original assignee, Dainippon Screen Mfg. Co., Ltd. (reel/frame 004388/0817), with no recorded transfers to a monetization vehicle — further confirming this patent was never an assertion asset. The maintenance-fee history tells the same story: fees paid at year 4 (1990-10-31) and year 8 (1995-01-03), then a reminder mailed 1999-02-09, lapse recorded 1999-07-18, and expiration effective 1999-07-21. The owner walked away.

Recommended next steps

If you are a defendant and someone asserts this patent:

1. Lead with expiry, not invalidity. The patent expired 1999-07-21 for failure to pay maintenance fees. Damages for the entire enforceable period are time-barred (the § 286 six-year lookback from any 2026 complaint reaches back only to 2020, decades after lapse), and there is no injunctive relief available against an expired patent. Request the maintenance-fee and legal-status record from PatentCenter for application 06/716,686 and the STCH event dated 2018-01-31.
2. Demand the asserted claim numbers and the asserted product. If the demand cites claims that survive in some other family member, note that the family members — JP JPH0685556B2, DE 3510825 C2, GB 2157120 B — are separate national rights with their own terms and their own expiry dates; none of them revives US 4,682,243.
3. There is no FWD to quote. I will not invent one. Any motion or letter arguing "claims 1–5 were canceled in IPR" would be fabricating a record; the only truthful statement is that claims 1–9 are untested and expired.
4. Verify the patent number on the face of any demand letter. Given the 4,682,243 / 7,682,243 collision described above, confirm whether the counterparty actually means the GAT gaming patent — in which case a genuine IPR record exists and the analysis changes completely.

If you are tracking this patent defensively (portfolio or FTO):

• Treat it as dead and non-blocking for any product shipping today. No watch-and-clear action, no design-around, and no IPR budget is warranted.
• Do not commission an IPR. It would be a waste of fees, and the Board has declined trials on expiry-adjacent grounds and discretionary factors where no live controversy exists; more fundamentally, there is no assertion to defend against.
• If your concern is the technical subject matter (constant-sampling-frequency magnification conversion with pixel averaging to suppress moire), the relevant live art is the citing family — e.g., US 4,809,083 (Sharp, "enlarging and reducing function") and US 4,812,915 (Minolta, "eliminates moire patterns") — and the modern equivalents in the H04N1/3935 and G06T3/40 classifications. US 4,682,243 is a historical citation node, not a live risk.

If you need authoritative confirmation of the "zero proceedings" finding:

• PTAB E2E / PTABX — search by patent number 4682243 in AIA trial proceedings.
• USPTO Open Data Portal — the source of the structured block.
• USPTO PatentCenter — full file wrapper for 06/716,686, to check for any ex parte reexamination certificate or adverse decision not captured in the ODP trial dataset.
• CourtListener — for any Federal Circuit docket, which would be empty absent an underlying Board decision.

Caveat, restated plainly: My conclusion is "no PTAB activity found across the canonical ODP list and independent searching." It is not a certification from PTAB E2E. The prior section flagged the same limitation for the CAFC check, and the same caveat applies here — the two sections are consistent, and neither was run against an authenticated docket system.

I'll research the assignment record for this patent, starting with the USPTO Assignment Center and related sources.

Let me check the Assignment Center interface and Dainippon Screen's corporate status.

Let me do a couple of final targeted checks on the assignment correspondent and any post-issuance transfers.

I've gathered what's available. Note a constraint up front: my tool searches reached Google Patents (the authoritative full text, including its legal-events mirror of the USPTO assignment record) and general web sources, but I could not run a live, authenticated query against USPTO Assignment Center, PEDS/ODP, or RPX/Unified directories. Findings below are grounded in the reel/frame and legal events reflected in the patent record, and I flag explicitly where I could not confirm something.

Inventors

Pattern check: Only one inventor, no co-inventor departures, no 12-month departure cluster. Nothing anomalous. There is no evidence in the record of Hatayama leaving or of any subsequent inventor-side activity. This is a routine corporate employee assignment, not a red flag.

Original assignee

Dainippon Screen Mfg. Co., Ltd. (Kyoto, Japan) — original and current assignee of record.

• Line of business: Graphic arts / prepress imaging equipment and, later, semiconductor and FPD production equipment. This patent sits squarely in their electronic image-reproduction / scanner-recorder business line (the "画像情報処理機器" / image-information-processing equipment division they expanded in the early 1980s). Dainippon Screen was a genuine, large operating company — a long-time Tokyo/Osaka Stock Exchange listed manufacturer (listed Osaka 2nd section 1962, Tokyo 2nd section 1967, moved to 1st section 1970; Prime Market from 2022).
• Did they ship a product embodying the claims? Highly likely on the record available — the specification is written as an improvement to their own commercial drum scanner/recorder systems (input drum, recording drum, multiple recording beams, helical scanning), and it distinguishes prior art assigned to Dainippon Screen itself (US 4,327,380; US 4,454,537; US 4,600,567). I could not independently confirm a specific product SKU. Unclear on the exact model, but there is no indication they were a paper-only patentee.
• Current status: Operating. The entity renamed itself SCREEN Holdings Co., Ltd. on 2014-10-01 as it moved to a holding-company structure; the graphic-arts and precision businesses were split into SCREEN Graphic Solutions Co., Ltd. (and later SCREEN PE Solutions). It remains a going concern (TSE Prime, code 7735) — not acquired, dissolved, or in bankruptcy.

Assignment timeline

Only one assignment is recorded against this patent.

• 1985-03-15 (executed) / recorded 1985-03-27 — Reel 004388 / Frame 0817
  • Conveyance: Assignment (recorded as "ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: Hatayama, Fumihiro
  • Assignee: Dainippon Screen Mfg. Co., Ltd., Kyoto, Japan
  • Correspondent: Not determinable from the sources I could reach. A 1985-vintage reel/frame recording predates the modern online correspondent field; the record available to me does not expose an attorney-of-record name, firm, or address for this entry. I will not fabricate one.
  • Context: Original corporate assignment — inventor-to-employer transfer at filing. Not a fire-sale, reorg, securitization, or transfer-to-asserter.
  • Recurrence flag: N/A — a single entry, no recurring correspondent.

Post-issuance transfers: None recorded. The Google Patents legal-event log for US 4,682,243 shows only: the 1985 assignment, maintenance-fee payments (4-year, 8-year, plus fee-payment procedures into 1995), a 1999 maintenance-fee reminder, and then lapse. There is no Merger, Change of Name, Security Agreement, License, Release, or Correction recorded against the patent — and notably, no assignment recording the 2014 Dainippon Screen → SCREEN Holdings / SCREEN Graphic Solutions reorganization. If the patent had still been alive and valuable in 2014, one would expect a re-recorded assignment for that internal split; its absence is consistent with the patent having been abandoned long before.

Timeline diagram

timeline
    title Ownership of US 4682243
    1984 : JP priority application filed
    1985 : Inventor assigns to Dainippon Screen
         : US application filed
    1987 : US patent issued
    1995 : Final maintenance fee paid
    1999 : Lapsed for nonpayment
         : Rights enter public domain
    2014 : Dainippon Screen renamed SCREEN Holdings
         : No patent assignment recorded

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. No operating-to-licensing-LLC transfer exists in the record. The only assignee ever recorded is Dainippon Screen itself (Reel 004388/0817).
2. Known asserter in the chain — Not present. No assignee matches any public NPE list (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Round Rock, etc.). The sole assignee is a listed Japanese capital-equipment manufacturer.
3. Repeat correspondent across the chain — Not present / unclear. Only one recorded assignment exists, so recurrence is impossible. (Correspondent identity itself is undeterminable from my sources — see the timeline entry.)
4. Cascading transfers — Not present. Zero post-issuance transfers; nothing to cascade.
5. Pre-litigation transfer — Not present. No assignment in the record precedes any lawsuit; in fact no lawsuit naming this patent surfaced in any search.
6. Bankruptcy fire-sale — Not present. Dainippon Screen never filed bankruptcy; it remains a solvent, listed operating company.
7. Privateering — Not present. No transfer to any third party, let alone an entity asserting on Dainippon Screen's behalf.
8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified, or OIN. It terminates where it began: with the original operating-company assignee, followed by abandonment.

Litigation: No Federal Circuit, district-court, or PTAB proceeding naming US 4,682,243 was surfaced by any search — consistent with the earlier CAFC-docket check in the summary section. A patent that lapsed in 1999 (as this one did) is essentially never asserted afterward.

Verdict

Insufficient data — more precisely, no NPE activity; single-assignment chain only.

The entire ownership record consists of one 1985 employee-to-employer assignment to Dainippon Screen Mfg. Co., Ltd. (Reel 004388/0817, executed 1985-03-15, recorded 1985-03-27). There are no post-issuance assignments, no shell-transfer, no asserter, and no litigation — the patent lapsed for non-payment of maintenance fees effective 1999-07-21, roughly 12 years after issuance, and no transfer was ever recorded even for the 2014 corporate renaming to SCREEN Holdings. This is the textbook "original assignee still owned it, then abandoned it" profile, which is the inverse of an NPE pattern.

Caveat: I could not directly query USPTO Assignment Center / PEDS or the RPX/Unified directories from these tools, so this is a "no records found beyond the original assignment" result grounded in the patent's own legal-event log, not a certified title search. To verify the reel/frame and to capture the 1985 correspondent of record, query: https://assignmentcenter.uspto.gov/ (search "4682243") — and note the record is also mirrored at https://assignment.uspto.gov/patent/index.html.

I'll research the prior art citations for US 4,682,243. Let me search the key references.

Let me research the remaining citations, especially the ones most likely relevant to the averaging/moire aspects.

Prior-art analysis for US 4,682,243

Source note / caveat. My retrieval reached Google Patents (which mirrors the USPTO front-page citation data) and Justia/uspto.report, plus general web results. I did not open USPTO PatentCenter or the paper file wrapper directly, and I could not pull full text for every cited document. The citation list below is taken from the front page of US 4,682,243 exactly as issued and indexed. Where I describe a reference's disclosure from its title/subject matter rather than from text I actually read, I flag it as inferred. No identifiers were auto-corrected.

1. The citation record on US 4,682,243's face

The patent lists two distinct groups:

"Patent Citations (9)" — cited by the examiner:

"Family Cites Families (3)" — cited in the corresponding foreign prosecution (the JP/DE/GB siblings):

All nine examiner citations and all three family citations have publication dates before the 1984-03-30 priority date, so each is facially available as prior art under pre-AIA § 102(a)/(b). (C2 issued 1986 but has a 1977 effective filing chain through abandoned Ser. Nos. 933,714 and 170,127, which the patent itself identifies — so it is § 102(e) prior art as well.)

Reminder of what must be met for § 102 anticipation. The independent claims are narrow in exactly one respect: the combination of (i) a sampling frequency held constant independent of the magnification ratio, (ii) averaging the digitized data over a number of pixels that corresponds to the magnification ratio, and (iii) writing at a frequency tied to the magnification ratio (f_in = K₂·M·f_s) while reading at a constant frequency. To anticipate claim 1, a single reference must teach all of (i)–(iii). To anticipate claim 3/6, all of the recited means (including the averaging means) must be present in one reference.

2. Reference-by-reference assessment

C1 — US 4,163,605 (Yamada), 1979-08-07, "Method of picture reproduction with variable reproduction scale."
Disclosure: Scans an original, A/D-converts image signals, writes them to a memory in synchronism with timing pulses g1, and reads them out under an address generator. Magnification/reduction is achieved by inserting "insert pulses" g3 between the regular timing pulses, without changing the base frequency of the timing pulses; the insert pulses cause addresses to be repeated (enlargement — same value written to two successive addresses) or skipped (reduction — a value is not converted and is erased). Claim 1 is the insert-pulse addressing method; claims 6–7 are the enlarge/reduce variants; claim 8 is the apparatus. The stated magnification M = g2/g2′.
Anticipation: Relevant to the general architecture of claims 1, 2, 3, 6 (scan → A/D → memory → read at base rate). But it does not teach constant sampling frequency tied to a magnification-dependent write frequency, and it does not teach averaging pixels. Its reduction mechanism discards pixels rather than averaging them. It therefore does not anticipate any independent claim; it is a background/§ 103 reference directed to the same "memory + variable scale" field.

C2 — US 4,605,957 (Dainippon Screen Seizo), 1986-08-12, "Machine and method for picture reproduction with steadily variable reproduction scale."
Disclosure: The patent's own specification discusses this "third method" family: all sampling pulses (first through fourth) are at the same frequency, at differing phases, and scale is set by a non-decreasing addressing scheme in which addresses are repeated (≥2 read-outs of one location) for enlargement or advanced by ≥2 (a location skipped) for reduction. The background portion expressly criticizes the earlier "vary the write/read frequency ratio" method for requiring wide-band A/D-D/A converters.
Anticipation: Same in-family as C1. It addresses the same problem (variable scale via memory addressing) but, notably, moves away from varying the frequency ratio and does not average pixels. Not an anticipatory reference for claims 1–9; useful for showing the state of the art and for § 103 motivation.

C3 — US 4,305,093 (International Electronic Photo Process Laboratory), 1981-12-08, "Method of producing multiple images in a scanning apparatus."
Disclosure: Divides a store into equal storage sections (one per color-separation/image to be produced), writes multiple digital color-separation signals in parallel into the sections synchronously with input sampling pulses ISP, then reads the entire store back serially in time-sharing using output sampling pulses OSP for a single-line output.
Anticipation: Concerned with multi-image/parallel memory handling, not magnification–ratio averaging. It says nothing about a constant sampling frequency chosen for a given magnification, nothing about averaging a magnification-dependent pixel count, and nothing about f_in = K₂·M·f_s. It is a peripheral § 102/§ 103 reference at best (possibly cited for its memory write/read sampling-pulse architecture). Does not anticipate claims 1–9.

C4 — US 4,327,380 (Dainippon Screen Seizo), 1982-04-27, "Control method for reproduction picture positions in a direction of a cylinder's axis."
Disclosure (inferred from title/assignee/date; I did not read the text): Sub-scanning (axial-direction) position/registration control for reproduction picture placement on the recording cylinder — i.e., the sub-scanning counterpart to the main-scanning magnification handling discussed at the end of the '243 specification.
Anticipation: Directed to picture positioning, not to sampling-frequency/averaging magnification conversion. No anticipation of claims 1–9.

C5 — US 4,454,537 (Dainippon Screen Seizo), 1984-06-12, "Method for reproducing a color picture for use in a picture reproducing machine."
Disclosure (inferred): Color-picture reproduction methodology (color scanning/recording), again in the same assignee's family.
Anticipation: No teaching of the constant-sampling + magnification-dependent averaging + magnification-dependent write-frequency combination. No anticipation.

C6 — US 4,470,074 (Dainippon Screen Mfg.), 1984-09-04, "Method of scanning and recording images."
Disclosure (inferred): Scanning/recording method — likely involving input beam/recording-beam control and aperture considerations of the kind the '243 specification discusses (beam diameter vs. aperture trade-off). Publication date 1984-09-04 falls after the 1984-03-30 priority date, but its U.S. filing (1981-05-27 priority) makes it § 102(e) prior art regardless.
Anticipation: Even if it touches aperture/beam control, it does not disclose sampling at a constant frequency plus averaging over a magnification-dependent pixel count plus a magnification-dependent write frequency. No anticipation; possibly relevant to the "aperture size control" problem statement.

C7 — US 4,521,812 (Dr.-Ing. Rudolf Hell GmbH), 1985-06-04, "Method of improving contrast accentuation."
Disclosure: Contrast intensification / unsharp-masking for electronic image processing (generating outer-field differential signals and superimposing a correction signal on the image signal).
Anticipation: This is an image-enhancement reference with no bearing on sampling frequency, magnification-ratio averaging, or memory write/read frequency relations. Not anticipatory for any claim of '243.

C8 — EP 0 085 351 A2 (Toshiba), 1983-08-10, "Picture information display device."
Disclosure (inferred): Display-side picture information handling. Display context rather than scanner/plotter magnification.
Anticipation: No teaching of the claimed combination. Not anticipatory.

C9 — EP 0 095 514 A1 (Dr.-Ing. Rudolf Hell GmbH), 1983-12-07, "Method and apparatus for the reproduction of an image with a coarser resolution than during picture scanning."
Disclosure (from the EP/DE family text): To convert a fine scanned image into a coarser (overview) image, it selects a number of pixels from the fine-image digital signals that correspond to one coarse image point, where the number "results from the scale" (e.g., a 4×4 field → one coarse pixel). The image signals of those pixels are weighted (individually, e.g., by a triangular or constant factor) and summed to form a single representative value that is stored in a coarse-image memory (memory 7). The hardware is expressly an adder (13) with a feedback register (14) plus a multiplier (12), and the "error" from summing is corrected by simply taking only the more-significant bit lines — i.e., an implicit division/shift for averaging. The field size (number of fine pixels per coarse pixel) can be "any other value" when a different scale factor is required.
Anticipation — this is the single most material reference:

• For claim 1 step (c) and claim 3 element (d) / claim 6 element (c) ("averaging the digitized image data among a number of pixels corresponding to the magnification ratio"), C9 discloses the summing/averaging of a scale-determined number of pixels into one stored value, using an adder + register + memory — structurally very close to the FIG. 3 (divider) and FIG. 5 (shift-register) averaging circuits of '243. The MSB-truncation technique it describes also parallels '243's ÷2ʳ shift-register approach.
• However, C9's stated purpose is producing a monitor/overview image at coarser resolution (a display layout aid), and it does not disclose (i) sampling at a constant frequency independent of the magnification ratio in the '243 sense, (ii) the write-pulse relation f_in = K₂·M·f_s, or (iii) reading out at a constant read frequency for recording a variable-magnification reproduction.
• Conclusion: C9 does not anticipate any independent claim in full, but it is the closest and most dangerous art for the averaging element and would be the lead § 103 reference (alone or combined with C1/C2) against claims 1, 3, 4, 6, 7.

F1 — US 4,233,636 (Eocom Corporation), 1980-11-11, "Optical read-write system with electronically variable image size."
Disclosure (inferred): Optical read/write (scanning/recording) system whose image size is varied electronically rather than optically. Relevant to the general "electronic magnification conversion" field.
Anticipation: No teaching of sampling at constant frequency + magnification-dependent averaging + f_in = K₂·M·f_s. Not anticipatory; background.

F2 — JP S57-162571 A (Matsushita), 1982-10-10, "Picture signal processing system."
Disclosure (inferred from title): Picture-signal processing. Without the full text I cannot attribute a specific disclosure; on the indexed record it is a general signal-processing reference.
Anticipation: Not shown to disclose the claimed combination. Not anticipatory.

F3 — JP S58-13066 A (Fuji Xerox), 1983-01-25, "Picture size contracting system."
Disclosure (inferred): Picture-size contraction — i.e., reduction. Reduction is exactly the direction in which '243's averaging is most relevant, so this is the family citation most worth pulling in full text; but contraction can be achieved by pixel decimation (as in C1) as well as by averaging, and the title alone does not establish averaging plus the two frequency constraints.
Anticipation: Not established as anticipatory for any claim; a candidate § 103 reference pending full text.

3. Bottom line

• No cited reference appears to anticipate any independent claim (1, 2, 3, 6) of US 4,682,243, because none of them discloses the three-element combination central to the invention: (i) sampling at a frequency held constant independent of magnification, (ii) averaging over a magnification-determined pixel count, and (iii) writing at f_in = K₂·M·f_s while reading at a constant frequency. The citations are overwhelmingly directed to the two competing approaches the '243 specification itself distinguishes (frequency-ratio variation, as in the background art; and unity-frequency insert-address/repeat-skip methods, as in C1/C2).
• Closest reference overall: C9 (EP 0 095 514, Hell). It is the only citation that expressly teaches summing/averaging a scale-determined number of pixels into a single stored value with an adder/register/memory and MSB-truncation, making it the primary § 103 reference against the "averaging" limitations of claims 1(c), 3(d), 4, 6(c) and 7 — but it lacks the constant-sampling and write/read-frequency relations, so it cannot alone anticipate.
• Secondary § 103 references: C1 (US 4,163,605) and C2 (US 4,605,957) for the memory-based variable-magnification architecture and for the "same base frequency" teaching, and C6/C4/C5 for the scanner/recorder environment. C3, C7, C8, F1–F3 are peripheral or non-analogous to the magnification/sampling-frequency problem.
• Dependent-claim notes: claim 4's divider-type averaging circuit (adder 22, register 23, data selector 24, preset counter 26, divider 25) is structurally akin to C9's adder/register/feedback-sum arrangement; claim 7's shift-register ÷2ʳ averaging likewise parallels C9's MSB-truncation. These are the claims most exposed by C9.

Caveats I want to be explicit about: (a) I verified C1, C2, C3, C9 against retrieved full text; my descriptions of C4, C5, C6, C7, C8, F1, F2, F3 are based on their front-page titles/assignees/dates plus subject-matter inference, not on text I read, so treat those as provisional; (b) I did not obtain authenticated USPTO PatentCenter output or the original examiner's Form PTO-892, so I cannot confirm the examiner's stated reasons for citation or whether any reference was applied under § 102 vs. § 103; (c) the analysis assumes pre-AIA §§ 102/103 given the 1984 priority date. If you need court-grade certainty on any single reference's disclosure, the next step is to pull the full text of C9 (and F3) and compare element-by-element against claim 1.

I'll ground my obviousness analysis in the actual prior art listed on the patent page. Let me pull details on the most relevant references.

Let me characterize the remaining cited references that could serve as secondary teachings.

Obviousness Analysis under 35 U.S.C. § 103 — U.S. Patent No. 4,682,243

Current date of analysis: April 26, 2026.
Scope note: This is a retrospective, hypothetical obviousness assessment. It builds on — and does not repeat — the claim summary and litigation findings already generated. All prior art is taken from the "Citations (9)" / "Family Cites Families (3)" / "Patent Citations" fields of the patent page, cross-checked against the references' own full texts where retrievable. Where I could not retrieve a reference's text, I say so.

1. Framework and assumptions

Applying Graham v. John Deere (scope/content of prior art; differences; PHOSITA level; secondary considerations) and KSR v. Teleflex (predictable combinations; "known technique to improve similar devices"; design incentives), the analysis below shows the independent claims are reasonably likely obvious over two- and three-reference combinations already on the face of the patent.

Important structural point: the nine references in the "Patent Citations" field are marked "Cited by examiner." They were before the examiner, and the claims nonetheless issued. That is meaningful — but it is not dispositive. A §103 case requires showing (i) each limitation is disclosed or suggested, and (ii) an articulated motivation to combine. I supply both below, and I flag the strongest non-obviousness counterargument in §6.

2. The relevant prior art (as disclosed in the reference texts)

I could not retrieve EP 0085351 A2 (Toshiba) and the two family-cited JP documents; they are not relied upon.

3. Where each claim limitation is taught

Claim 1 (independent method):

Claim 2 / claim 6 (2ʳ-pixel averaging via shift/divide): Hell's bit-stripping ("remaining bits are stripped") is literally division of the summed signal by a power of two by truncating r low-order bits; the "shift register" of claim 7 and the "÷1/2ʳ frequency divider" of claim 9 are the standard hardware equivalents. The patent's own specification supplies the motivation: the divider of the FIG. 3 embodiment "is expensive, however, the following method may be employed" (the shift-register/2ʳ variant). That is an applicant admission that substituting a bit-shift for a divider is an obvious cost optimization.

Claim 3 / claim 6 (apparatus): same elements, in means-plus-function form.

Claim 4 / claim 7 (averaging circuit): Hell's FIG. 1 is an adder (13) + feedback register (14) + memory (7) accumulate-and-dump structure. The added "data selector applying zero on receipt of the write pulse" is simply resetting the accumulator at each group boundary — the routine counterpart of Hell's control-unit reset "at the last image point of the field." The preset counter (claim 4) is a standard sampling-pulse counter.

Claims 5, 8 (pulse signal distributor): the patent itself calls this a component "having a well-known construction." Claim 9 (÷1/2ʳ divider) is routine and equivalent to Hell's bit-stripping.

4. Combinations that render the claims obvious

Ground A — Hell + Eocom (claims 1 and 3; best two-reference case)

• Hell supplies the core step (c): reduce data volume by averaging N fine pixels, N set by the scale factor, and correcting magnitude by bit-shifting.
• Eocom supplies steps (b)+(d): hold the input sampling clock constant and vary the memory write clock in proportion to the desired size change, achieving magnification by the clock differential.
• Result: constant sampling frequency + averaging by scale + write-clock proportional to magnification = claim 1. Substitute means for steps in a memory scanner = claim 3.

Ground B — Hell + Yamada + US4605957 (claims 1–3, 6; adds same-assignee art)

• US4605957 (same assignee; named in the patent's own Background) supplies the memory-based variable-scale scanner with constant sampling frequency.
• Yamada supplies the explicit magnification relation M = write/read address ratio, i.e., a write clock tied to M.
• Hell supplies the averaging-over-N-pixels and bit-shift average.
• Together these read on claims 1, 2, 3, 6 and the dependent details.

Ground C — Hell alone as primary, with routine design choices (claims 2, 6, 7, 9)

Hell already discloses (i) averaging a scale-dependent number of pixels, and (ii) correcting the too-large sum by stripping low-order bits — i.e., dividing by a power of two via a shift. Claim 2's "2ʳ / dividing by shifting" and claim 7's "shift register" are thus disclosed or, at minimum, an obvious hardware substitution expressly motivated by the specification's own cost rationale.

Dependent claims

Claims 4, 5, 7, 8, 9 are not separately patentable: Hell's adder/register/memory with bit-stripping (claim 7), a "well-known" pulse distributor (claims 5, 8), a preset counter, and a ÷1/2ʳ frequency divider are each conventional elements arranged to perform their known functions. Under KSR, "the combination of familiar elements according to known methods is likely to be obvious."

5. Motivation to combine (articulated under KSR)

1. Same field and same problem. All three primary references are electronic image scanning/recording systems confronted with changing image scale. Hell's stated object is to reproduce an image "with a coarser resolution … without a greater loss of information" — which is exactly the information loss that produces moire when sampling is reduced in US4682243.
2. The patent admits the problem and a known solution. Its Background Art states moire arises "when the [sampling frequency] is substantially lower than the [reading] frequency," and that "a process has recently been introduced for reducing moire by varying the diameter of an input scanning beam." Recognizing a known problem with a known (if awkward) solution, and substituting a better-known technique, is the paradigm of KSR obviousness.
3. Finite, predictable set of solutions. The Background identifies only three scaling routes (vary drum speed; vary write/read clock ratio; skip/repeat addresses). Averaging to preserve information (Hell) applied to the frequency-ratio route (Yamada/Eocom) is a predictable combination of known options.
4. Design incentive / known drawback of the prior route. US4605957 expressly warns that varying sampling frequency requires "a special A/D converter and memory having wide frequency range." A PHOSITA is therefore motivated to keep sampling frequency high/constant and vary the memory write clock instead — precisely the claimed architecture.
5. Reasonable expectation of success. Both the averaging (Hell) and the memory clocking (Yamada/Eocom/US4605957) were proven, off-the-shelf techniques; expecting the aggregate to scale images with reduced moire is reasonable without experimentation.

6. The strongest non-obviousness arguments (and my assessment)

7. Bottom line

• Claim 1 is the most exposed: Hell (EP0095514A1 / US4533942) in view of Eocom (US4233636) discloses or suggests every step; Yamada (US4163605) and US4605957 reinforce the write-clock-to-magnification relation and the constant-sampling architecture.
• Claim 3 falls with claim 1 (same elements, apparatus form).
• Claims 2, 6, 7, 9 are additionally rendered obvious by Hell's own bit-stripping and by the specification's admission that the shift-register/2ʳ variant is a cheaper substitute for the divider.
• Claims 4, 5, 8 recite conventional circuit components arranged to perform their known functions.

Net assessment: On the references appearing on the face of US 4,682,243, there is a strong prima facie case of obviousness under § 103 for the independent claims and most dependents, most cleanly via Hell + Eocom (claims 1/3) augmented by Yamada and US4605957 (claims 2/6–9).

Caveats (read before relying on this):

1. This is a documentary opinion, not a claim chart from the prosecution record. I could not retrieve the JP priority document (JP 59‑64657), the JP publications JP 54‑65601 / S57‑162571 / S58‑13066, or EP 0085351 A2 (Toshiba); none is relied upon.
2. The EP 0095514 A1 text I used is a machine-translated summary; the US counterpart US 4,533,942 was used to confirm the adder/register/bit-stripping disclosure. Quotation wording from the EP summary should be verified against the certified text before citing in a legal filing.
3. No secondary-considerations evidence was available on the patent page; if such evidence exists elsewhere, the conclusion could change.
4. The patent is expired (lapsed 1999), so this §103 analysis is of historical/academic interest rather than current enforceability.