US 6765616B1

I'll search for authoritative information on this specific patent number.

Let me search for litigation and any 2026 appellate activity specific to this patent.

One more targeted search on appellate dockets for this patent.

US 6,765,616 B1 — "Electric Camera"

Search note up front: I ran targeted queries against the web (Google Patents, USPTO/PTAB filings, uspto.report, court dockets). I did not find any 2026 Court of Appeals for the Federal Circuit docket in which US 6,765,616 (cited as "6765616B1") is itself the patent on appeal. I did surface 2026 Federal Circuit Maxell-related dockets (e.g., [Samsung Electronics Co.](/litigations/by-defendant/Samsung%20Electronics%20Co.) v. Maxell, No. 26-1320, and Maxell, Ltd. v. Samsung Electronics Co., No. 26-1330), but I could not confirm that any of those specifically involve the '616 patent. Treat any 2026 appellate connection as unverified. I have flagged uncertainties explicitly below.

Bibliographic data (from Google Patents record)

Ownership chain recorded on the face of the record: Hitachi, Ltd. → Hitachi Consumer Electronics Co., Ltd. (2013) → Hitachi Maxell, Ltd. (2014) → Maxell, Ltd. (2018) → Maxell Holdings, Ltd. (2021 merger) → Maxell, Ltd. (2021 name change).

Source: https://patents.google.com/patent/US6765616/en

Family / continuation lineage (from US 9,544,517 B2, which claims priority as a continuation of the '616 application): 09/520,836 → US 8,059,177 (div. 10/660,710); 10/660,711 → US 7,403,226 → US 8,339,493 (12/845,266) → US 8,736,729 (13/681,495) → US 9,100,604 (14/264,243) → US 9,544,517 (14/661,227) → US 10,554,917 (15/386,656); and US 10,462,403 (15/959,774). This means the '616 patent sits at the root of a substantial continuation family.

Abstract (verbatim)

"A photography related to video cameras, camcorders, digital still cameras and others using a solid-state image sensing device, and particularly an electric camera using a solid-state image sensing device with a large number of pixels. The use of an image sensing device with a sufficient number of pixels for still image photographing ensures good performance for the moving image photographing and for the monitoring of a static image photographing. The image sensing device used has an arbitrary number of vertically arranged pixels equal to or more than three times the number of effective scanning lines of the television system. During the moving image photographing and during the monitoring of a static image photographing, the pixels are cyclically mixed together or culled and desired television signals are generated by performing the signal reading processing during the vertical blanking periods and the interpolation processing. During the static image recording, signals of all pixels are read out independently and recorded."

Plain-language overview of the independent claims

Claim 1 — the core camera claim. It covers an electric camera built around a vertical-stripe color filter (three colors repeating horizontally every three pixels, same color down each column) and an image sensor whose vertical pixel count N is at least three times the effective scanning-line count M of the television display screen. A driver circuit makes the sensor vertically mix or cull the accumulated signal charge of every K pixels so that the number of output lines matches M; K can be any integer ≤ ⌊N/M⌋. A signal processing unit turns those sensor outputs into image signals, and there is a display unit with the TV screen. The key inventive hook is the last three limitations: the driver uses a different mixing/culling factor K1 for a first set of images and K2 for a second set of images (K1 ≠ K2) — i.e., reconfigurable vertical resolution depending on what is being shot/displayed. (In the spec this maps to 4-pixel mixing for moving video and 5-pixel mixing for still-image monitoring, from a 1200-line sensor.)

Claim 13 — the same concept, differently framed. This independent claim recites the sensor as a grid-arrayed pixel array with an arbitrary number N of vertically arranged pixels (N ≥ 3M) and recites the color filters as arranged on the grid pixels in the repeating vertical-stripe pattern, then recites the same driver (mix/cull every K, K ≤ ⌊N/M⌋), signal processing unit, display unit, and the same K1 ≠ K2 conditional. Functionally it is a parallel/companion rendition of claim 1 with the color-filter/array language restructured.

Notable dependent claims (claims 2–12 depend from claim 1; 14–20 from claim 13)

• Claim 2 / 14: The driver reads out (or discards) the N − K·M excess pixel rows during the vertical blanking period, extracting only the K·M rows during the vertical effective scanning period to get M output lines.
• Claim 3 / 15: The signal processor extracts a horizontal segment matching the TV aspect ratio and expands it across the whole horizontal effective scanning period (aspect-ratio correction).
• Claim 4 / 16: The driver shifts which pixels are mixed/culled each display cycle to produce interlaced output.
• Claim 5 / 17: The processor interpolates the vertical gravity centers of the interlaced signals so the two fields are exactly 180° out of phase.
• Claim 6 / 18: Ties the two image types to moving images (K1) vs. still images (K2), with K1 < K2 (coarser mixing for motion, finer for stills).
• Claim 7: Adds a gain-adjust unit whose gain changes with the number of pixels mixed so the output level stays constant.
• Claim 8: Adds an image-unstability (shake) detector; the size and position of the extracted area are shifted to correct for camera shake.
• Claim 9: Adds a trigger device (e.g., shutter button); on trigger, no mixing occurs and all pixels are read out independently.
• Claims 10–12: The three filter colors are yellow/green/cyan, yellow/white/cyan, or red/green/blue respectively.

Caveat: the claim text available to me ends mid‑claim 18 ("…the value of K 1 is smaller tha…"), so claims 19–20 are not fully reproduced; I am not characterizing them.

Litigation / PTAB context (verified in search results)

• IPR2019-00087, BlackBerry Corporation v. Maxell Ltd., Patent 6,765,616 B1 — filed by BlackBerry; Google Patents lists the outcome as Settlement. It was one of a batch of five BlackBerry IPRs (‑00087 '616, ‑00088 '334, ‑00089 '394, ‑00095 '897, ‑00097 '604). Source: PTAB decision granting pro hac vice admission (docketalarm / PTAB records).
• Delaware District Court cases involving the '616 patent as asserted/counter-asserted patent: 1:17-cv-01446, 1:18-cv-00216, 1:18-cv-00310, plus E.D. Tex. 5:18-cv-00136 (per Google Patents litigation links and the D. Del. opinion at 18-216). The D. Del. opinion confirms the '616 was among patents counterclaimed by a defendant.
• The '616 patent is cited as a prior-art reference in later Maxell/Hitachi patents (e.g., listed in the "References Cited" of US 9,544,517 B2 and US 10,462,403) and appears in terminal disclaimers over later continuation applications (e.g., Application 14/264,243).

Explicit uncertainties

1. No confirmed 2026 CAFC docket for the '616 patent. The 2026 Federal Circuit Maxell dockets I located (26-1320, 26-1330) are unverified as to whether '616 is among the patents on appeal. If you need appellate certainty, that should be checked directly in PACER/CM-ECF for the Federal Circuit.
2. Current assignee is listed by Google Patents as Maxell, Ltd. with the caveat that Google states assignee data "may be inaccurate" and is not a legal conclusion.
3. Legal status ("Expired – Lifetime," anticipated expiration 2020-03-08) is Google Patents' characterization, not an official USPTO determination.
4. Claims 19–20 were not available in full in the retrieved text; I have not summarized them.
5. I am relying on the authoritative full patent text provided (Google Patents) over any conflicting search snippets, per your instruction; where search results and the patent text agree (title, inventors, dates, assignee chain), I have so noted.

I'll research litigation involving US patent 6765616B1. Let me search multiple sources.

I found several leads. Let me dig deeper into the specific cases.

Let me verify which cases specifically asserted the '616 patent and check the PTAB proceedings.

Litigation Involving US Patent 6765616B1 ("Electric Camera")

Patent identification (as given): US 6765616B1 — "Electric Camera." Inventors: Takahiro Nakano, Ryuji Nishimura, Toshiro Kinugasa. Original assignee: Hitachi, Ltd.; current assignee of record: Maxell, Ltd. (successor via Hitachi Consumer Electronics → Hitachi Maxell → Maxell). Priority 2000‑01‑11; filed 2000‑03‑08; granted 2004‑07‑20; anticipated expiration 2020‑03‑08 (status: Expired – Lifetime).

Note on scope: This patent number appears in the public record as a 7‑digit number (6,765,616). All cases below are tied to that literal number as it appears in the cited sources. I have not auto-corrected or normalized any identifier. Where a source labeled the case only generically, I flag the limitation.

District court litigation

1. Maxell, Ltd. v. BlackBerry Corporation and BlackBerry Ltd.

Sources: Unified Patents litigation portal, https://portal.unifiedpatents.com/litigation/Delaware%20District%20Court/case/1:17-cv-01446 ; Stanford NPE Litigation Database, https://npe.law.stanford.edu/patent/6765616-0 ; Docket Alarm, https://www.docketalarm.com/search/?q=patent:(7995897).

2. Olympus Corporation and Olympus America Inc. v. Maxell, Ltd. (declaratory judgment)

Sources: Justia Dockets, https://dockets.justia.com/docket/delaware/dedce/1:2018cv00216/64447 ; UniCourt, https://unicourt.com/case/pc-db1-olympus-corporation-et-al-v-maxell-ltd-[673988](/patent/673988).

3. Maxell, Ltd. v. Olympus Corporation and Olympus America Inc.

Sources: Justia Dockets, https://dockets.justia.com/docket/delaware/dedce/1:2018cv00310/64593 ; Docket Alarm, https://www.docketalarm.com/cases/Delaware_District_Court/1--18-cv-00310/.

4. Maxell, Ltd. v. ASUSTeK Computer Inc.

Sources: Docket Alarm, https://www.docketalarm.com/cases/Texas_Eastern_District_Court/5--18-cv-00136/ ; RPX Insight, https://insight.rpxcorp.com/litigation/txedce-[185754](/patent/185754)-maxell-v-asustek-computer.

PTAB proceeding (not district court, but relevant)

IPR2019‑00087 (Unified Patents — Petitioner)

Source: Google Patents litigation link, https://portal.unifiedpatents.com/ptab/case/IPR2019-00087.

Summary table

Confidence and limitations

• High confidence: The patent 6,765,616 B1 is identifiable in the Maxell v. BlackBerry (1:17‑cv‑01446) docket and in the Olympus declaratory‑judgment complaint (1:18‑cv‑00216), and it is tied to the ASUSTeK case and PTAB IPR2019‑00087 through Google Patents' litigation links.
• Lower confidence / not verified: The specific asserted count/claim set for the '616 patent in each case, and the final dispositions of the BlackBerry and Olympus matters. My search budget was exhausted before I could pull the final docket entries from PACER for those cases.
• Not the same patent: Several search results involve other Maxell/Hitachi patents in the same campaign (e.g., 7,199,821; 6,748,317; 8,339,493; 8,736,729; 8,098,695; and the Samsung patents 8,982,086; 10,176,848; 11,017,815). Those are different patents and are excluded here. I did not auto-substitute any of them for 6765616B1.

If you need PACER‑verified final judgments for the BlackBerry and Olympus cases, that would require a fresh docket pull (PACER or CourtListener RECAP) beyond what the retrievable sources above confirm.

I'll verify the PTAB record with targeted searches before writing.

I found a second, earlier potential proceeding (IPR2018-00904). Let me verify which patent it targets.

Now let me pin down the outcome of IPR2019-00087 and check for any other proceedings.

Proceedings overview

One AIA trial proceeding has been filed against US 6,765,616 B1 — IPR2019-00087, BlackBerry Corporation v. Maxell, Ltd. — and it terminated by settlement before the Board ever reached an institution decision. Breakdown: 1 settled / 0 claims invalidated / 0 claims sustained / 0 institution denials on the merits / 0 appeals. All 20 claims (1–20) remain UNTESTED at the PTAB.

Flagged contradiction with the structured input: the "PTAB proceedings on file" block states that the USPTO ODP API returns no AIA trial proceedings for this patent. That is inconsistent with the record I retrieved. Google Patents' own litigation block for this patent lists "PTAB case IPR2019-00087 filed (Settlement)," and third-party PTAB dockets (Docket Alarm, Patexia, the published pro hac vice order) confirm IPR2019-00087 is captioned against Patent 6,765,616 B1. Treat the ODP "no proceedings" result as a data-completeness gap, not as evidence of no activity — the proceeding is real, it just pre-dates or falls outside ODP's AIA ingest for this number (it was filed in October 2018 under the pre-ODP PTAB docketing system and terminated at the preliminary stage). I am following the web-sourced record here.

Bottom-line defensive posture: This is not a "hardened by IPR" patent and not a "claims already canceled" patent. It is a never-adjudicated patent. No petitioner has ever obtained a Final Written Decision on any claim of the '616 patent, so no claim is canceled and no claim is confirmed. Two practical consequences dominate: (1) a defendant gets no free § 315(e)(2) estoppel benefit from BlackBerry's settlement; and (2) any validity challenge is still a wide-open first bite at the apple. The counterweight is that the patent's statutory term ran out on 2020-03-08 (Google Patents' "Expired – Lifetime," anticipated expiration), so the realistic exposure today is backward-looking damages for pre-expiration conduct, not an injunction or ongoing royalty.

IPR2019-00087 — BlackBerry Corporation v. Maxell, Ltd.

• Type: Inter Partes Review
• Filed: 2018-10-15. Grounded in the reference lists of the later family members (US 10,462,403 / US 10,554,917), which cite "Maxell, Ltd., Petition for Inter Partes Review of U.S. Pat. No. 6,765,616, Inter Partes Review No. IPR2019-00087 dated Oct. 15, 2018"; Patexia's docketing profile for APJ Michael R. Zecher likewise lists the case as filed 2018-10-15.
• Status: Settlement (per Google Patents' litigation block). Plain-English gloss: the parties settled and the proceeding was terminated before institution; there is no Final Written Decision and no institution decision on the merits. Note: I could not confirm the exact termination date from the sources I retrieved — I am not going to guess it. The last confirmed activity I can date is the Board's 2019-01-23 order (Paper 9) granting Petitioner's motion for pro hac vice admission of Stephen McBride, which recites IPR2019-00087 among five consolidated BlackBerry/Maxell matters and shows the case was still pending on that date.
• Judge panel: APJ Terrence W. McMillin signed the pro hac vice order (Paper 9, 2019-01-23) covering IPR2019-00087 (along with IPR2019-00088, ‑00089, ‑00095, ‑00097). Patexia's judge profile for APJ Michael R. Zecher also lists IPR2019-00087 among his cases. Because the case never reached institution, no merits panel was ever designated — treat the panel attribution as administrative-only and not dispositive of anything.
• Petition grounds: Not retrieved. I could not obtain the IPR2019-00087 petition text, so I will not characterize which claims were challenged, which references were asserted, or whether the grounds ran under § 102 or § 103. What is confirmable from the docket is the caption and patent number only. (Do not rely on the sibling petitions' grounds as a proxy — the BlackBerry batch covered five different patents: '616, 6,973,334, 8,170,394, 7,995,897, and 9,100,604, each with its own art.)
• Institution decision: None issued. The Board granted the pro hac vice motion on 2019-01-23 but never decided whether to institute trial.
• Final Written Decision: None. The Board never reached the merits, so not a single claim of 6,765,616 B1 has been canceled or upheld in an AIA trial.
• Settlement / termination: The case resolved in connection with the broader Maxell–BlackBerry litigation settlement (BlackBerry was facing Maxell, Ltd. v. BlackBerry Corp., No. 1:17‑cv‑01446 (D. Del.), in which the '616 patent was asserted). Google Patents records the PTAB outcome as "Settlement"; contemporaneous reporting (Bloomberg, via Seeking Alpha, "BlackBerry settles Maxell patent lawsuit") confirms a global settlement of the BlackBerry/Maxell disputes. The terms are confidential — no license agreement or settlement amount is public in the materials I retrieved, and because there was no FWD the file contains no merits findings at all.
• Appeal: None possible on the merits — there was no FWD to appeal. No Federal Circuit docket on the '616 patent exists in the record I retrieved.
• Defensive value: Modest but real, and mostly negative-shaped for the patent owner. The favorable part: because the proceeding died pre-institution, no § 315(e)(2) estoppel attaches to BlackBerry or its privies, and — critically — no claim was narrowed by claim construction or disclaimer during the IPR, so the claim scope a defendant faces today is the unamended claim scope. The unfavorable part: BlackBerry's settlement gives you zero leverage from the PTAB record. You cannot tell a court "claims 1 and 13 were already invalidated" — they weren't. If you want validity leverage, you have to build it yourself from scratch; there is no borrowed FWD to lean on.

Sibling-proceeding context (NOT on the '616 patent — same specification, different claims)

These matter because they are the only place the Board has ever expressed a substantive view on the '616 patent's signature limitation, and they are the only estoppel-bearing events in the family. Do not cite these as proceedings on the '616 patent.

• IPR2020-00597 — Apple Inc. v. Maxell, Ltd. (U.S. Pat. No. 8,339,493) — institution DENIED, Paper 9, 2020-09-24. This is the single most useful data point for a defendant. The Board identified "a potential deficiency in Petitioner's evidence regarding the use of different intervals when mixing or culling signal charges accumulated in the vertically arranged pixel lines based on whether the electric camera is monitoring in a static image mode or recording in a moving video mode." That is, in substance, the K1 ≠ K2 limitation of '616 claims 1 and 13. The Board found Apple's art did not teach it. So the Board has taken the K1≠K2 limitation seriously — a defendant attacking claims 1/13 needs art squarely on the different-mixing-interval-per-mode point, not vague multi-mode cameras. (Quoted in Maxell's discretionary-denial briefing, IPR2025‑00392, Paper 7.)
• IPR2018-00904 — Olympus Corporation, Olympus Corporation of the Americas, and Olympus America Inc. v. Maxell, Ltd. (U.S. Pat. No. 8,339,493) — filed 2018-06-20, terminated by settlement 2018-12-26 (Paper 9), dismissed with prejudice as part of a Maxell–Olympus patent license agreement filed as confidential business information (Ex. 1018) under 35 U.S.C. § 317(b). Panel included APJs Lynne E. Pettigrew, Minn Chung, Terrence W. McMillin, and John A. [Hudalla]. Order: "These proceedings are at an early stage, and we have not yet decided whether to institute a trial." The '616 patent appeared here only as an exhibit (Ex. 1016, excerpt of the file history of application 09/520,836) because the '493 patent claims benefit of it — the '616 patent was not itself under challenge. The six-case Olympus batch (‑00904, ‑00906, ‑00908, ‑00909, ‑00910, ‑00911) also never reached institution.
• IPR2019-00068 — ASUSTeK Computer Inc. and ASUS Computer International v. Maxell, Ltd. (U.S. Pat. No. 7,403,226) — a '616-family patent (also claiming benefit of the 09/520,836 application). Joint motion to terminate under § 317 filed after a global settlement and stipulated dismissal of the parallel district court case; terminated without prejudice pre-institution. Every one of these family IPRs ends the same way: settle before institution.
• Ex parte reexamination 90/014,628 (U.S. Pat. No. 8,339,493) — not an AIA trial. Claims 5 and 6 confirmed; the § 103 rejections over Casio were withdrawn after Maxell antedated the Casio reference. Recorded in later petitions as an ex parte reexamination certificate (LG‑1005).

Pattern signal: across the '616 family, the Board has never once issued a Final Written Decision. Every AIA challenge — BlackBerry, Olympus, ASUS — ended either in settlement/termination pre-institution or (Apple, IPR2020‑00597) in a merits-based institution denial. Maxell uses IPRs as a leverage-resolution mechanism, not as a forum it litigates to judgment.

Strategic summary

Claim status of US 6,765,616 B1. CANCELED: none. SUSTAINED by PTAB: none. UNTESTED: all of claims 1–20. There is no IPR, PGR, or CBM Final Written Decision touching this patent, and there is no certificate of correction, disclaimer, or reexamination certificate narrowing any claim of it in the materials I reviewed. Whatever claim scope you face, you face the issued scope, unamended and unadjudicated. That cuts both ways: the patent has never been held valid by the Board, but it has never been narrowed either. Note also the term is expired (2020-03-08), so claims 1–20 are only reachable for pre-expiration conduct; a § 112 indefiniteness or § 101 challenge to a dead claim is generally a poor use of resources compared to a § 102/§ 103 attack on pre-2020 damages.

Estoppel landscape. This is the standout feature of the '616 dossier: there is essentially no estoppel lying around to inherit. § 315(e)(2) estoppel requires a final written decision; IPR2019-00087 settled pre-institution, so BlackBerry, BlackBerry Ltd., and their privies are not estopped — and, more importantly for you, you are not barred by anything BlackBerry did. No ground was "raised or reasonably could have been raised" in a trial that never began. Conversely, Maxell is not estopped from anything either (patent owners aren't subject to § 315(e) anyway). Practical upshot: every prior-art ground is available to you, including the § 102(e) Misawa '607 / JP H11‑355665, Parulski '218, Watanabe '236, Casio QV‑8000SX, Juen '162, Okino '947, and Matsuzaka '713 references that Olympus and Apple assembled in the sibling fights, plus the Watanabe '236 reference the examiner cited against the parent application and that the applicant allegedly never put in an IDS. The catch: § 325(d) discretion. Maxell has successfully attacked later family petitions for using art "previously presented to the Office" (see its § 325(d) briefing in IPR2025‑00392 on the '493 patent, arguing Misawa '607, Juen '162, Casio, Kurashige, and Parulski '218 were all already before the examiner/Board). Expect the same § 325(d) and General Plastic-style arguments against you. Pick art that is genuinely new to the Office or that you can tie to a material examiner error.

Pattern signals. (a) Single petitioner, single petition on the '616 patent — no repeat-filer pattern against this specific patent; the multi-petition filings (five BlackBerry IPRs, six Olympus IPRs, the Apple 2020 batch) were portfolio-wide campaigns against Maxell, not patent-specific. (b) No defensive aggregator. Unified Patents appears in the '616 record only as legal authority cited in the Board's pro hac vice order (Unified Patents, Inc. v. Parallel Iron, LLC, IPR2013‑00639) — that is a citation, not participation. The petitioner of record is BlackBerry Corporation, a commercial adversary defending a D. Del. suit, not a defensive aggregator. (c) Maxell does not appeal. With no family FWD, there is no Maxell PTAB appeal line; the "Maxell aggressively appeals PTAB losses" theory has no support in this family from the sources I retrieved. (d) Maxell's playbook is licensing. The Maxell IP Innovation Business Unit demand letters (e.g., to LG, 2021-07-19) and the succession of settlements (Olympus, ASUS, BlackBerry; Apple settled the E.D. Tex. case 2021-03 on the eve of trial) show that Maxell converts assertions into licenses. That is your realistic endgame, and the absence of any adverse PTAB judgment is precisely why Maxell can keep the '616 patent on its licensing sheets.

Explicit uncertainties. (1) The exact termination date of IPR2019-00087 and the claims challenged are not confirmed in my sources; I have not stated them. (2) APJ panel composition for IPR2019-00087 is administrative-signature evidence only. (3) Current assignee (Maxell, Ltd., via Maxell Holdings, Ltd.) is Google Patents' listing and carries Google's own accuracy caveat. (4) Reexamination 90/014,628 was against the '493 patent, not the '616. (5) Any 2026 Federal Circuit activity involving the '616 patent remains unverified — no such docket was located. (6) I have relied on the authoritative full patent text supplied in the prompt where it conflicts with search snippets; no such conflict arose as to the claims, which are unamended.

Recommended next steps

1. Do not represent to a court that any claim of the '616 patent has been invalidated at the PTAB. It has not. The only IPR, IPR2019-00087, settled pre-institution on the Board's 2019-01-23 pro hac vice order record; Google Patents labels it "Settlement." Pull the complete file via USPTO PTAB E2E (https://ptacts.uspto.gov/) to confirm the termination date and, if you need it, the challenged claims — that is the one gap in this analysis I could not close.
2. Lead with expiration. Term ran 2020-03-08 (anticipated). For a defendant facing a demand letter citing claims 1–20 in 2026, the threshold response is that the patent is expired and the only live theory is pre-2020 damages. Ask for the accused-conduct dates and the damages model before spending money on validity.
3. If you do file an IPR, attack claims 1 and 13 head-on at the K1 ≠ K2 limitation — and read the Board's IPR2020-00597 Decision Denying Institution, Paper 9 (2020-09-24) first. The Board found Apple's evidence deficient on "the use of different intervals when mixing or culling signal charges accumulated in the vertically arranged pixel lines based on whether the electric camera is monitoring in a static image mode or recording in a moving video mode." Art that shows a multi-mode camera is not enough; you need a reference teaching different vertical mixing/culling intervals for different modes. Given claim 6's K1 < K2 (motion vs. still) and the spec's explicit 4-pixel/5-pixel split, target that correspondence directly.
4. Budget for § 325(d) and discretionary denial. Maxell routinely argues that family art was "previously presented to the Office." Expect arguments that Misawa '607 (and its JP H11‑355665 priority), Juen '162, Casio, Parulski '218, and Takase '483 are duplicative. Build a § 325(d) Advanced Bionics step-two showing of material examiner error, and consider art wholly outside the family prosecution history.
5. Consider whether an IPR is even the right tool. The patent is expired; there is no injunctive risk and no ongoing-royalty risk. Money is better spent on (a) non-infringement on the color-filter limitation — note that in Olympus Corp. v. Maxell, Ltd., No. 1:18‑cv‑00216 (D. Del.), Olympus' entire '616 non-infringement theory was that its products "lack color filters to pass 'colors respectively arranged to cyclically appear horizontally at [an] interval and to pass the same colors arranged vertically," i.e., no vertical-stripe filter array; and (b) damages apportionment for pre-2020 sales.
6. Watch for new filings, but do not expect them. A 2020-expiry patent with no FWD is an unlikely IPR target. If a new petition does appear, note that trial must conclude within the statutory 1 year from institution (§ 316(a)(11)), with the institution decision due ~6 months from the § 315(b) service-date-triggered filing window.

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

Let me dig deeper into specific assignment records and reel/frame numbers.

I found reel/frame data for a sibling patent. Let me now verify the specific records for the '616 patent.

I have enough grounded material to answer. Note that my tool access terminated before I could pull the Assignment Center record directly, so I flag below exactly which reel/frame data is confirmed versus inferred.

Assignment & Ownership Analysis — US 6,765,616 B1 ("Electric camera")

Verification entry point: https://assignmentcenter.uspto.gov/ (search patent number 6765616). Cross-referenced: Google Patents legal events (page text supplied), Espacenet INPADOC, D. Del./E.D. Tex. complaints, PTAB IPR2019-00087, and Maxell corporate filings.

Retrieval caveat up front: My live queries returned the corporate transaction data and the reel/frame for a sibling patent, but I was not able to open the Assignment Center record for '616 itself to read the recorded correspondent field for each entry. Where I cite a reel/frame, it is confirmed for US 6,865,677 (a contemporaneous Hitachi bulk assignment) and only inferred for '616 — flagged inline. I did not fabricate reel/frame numbers for the entries I could not verify.

Inventors

Pattern note: All three inventors appear as assignors on the original assignment recorded 2000-03-08 (same date as filing). This is the standard Japanese corporate-inventor arrangement — rights pass to the employer at filing with no gap. There is no evidence of the "inventors depart within 12 months" fire-sale precursor: the inventors simply drop out of the record after 2000, and the patent stayed inside the Hitachi corporate family for 13 years (until 2013). Nakano is also the first-named inventor on the later continuation applications, which is consistent with continued employment, not a departure.

Original assignee

Hitachi, Ltd. (Japan) — named on the issued patent (2004-07-20).

• Line of business: diversified electronics conglomerate; the business unit relevant here was Hitachi's consumer video/camcorder (CCD) products — the patent is classified H04N (television/camera circuitry) and describes a CCD camcorder architecture.
• Did it ship a product embodying the claims? Yes — plausibly/typically. Hitachi sold CCD camcorders (the VM/Digital-MV lines) and digital still cameras during this era, whose sensor front-ends would read on the claim concepts (vertical pixel mixing/culling to TV line counts, K1≠K2 mode switching). I cannot, from the retrieved record, cite a specific model-number-level mapping — treat "shipped product" as highly likely but not documentary.
• Current status: Operating (still a global conglomerate). However, the IP at issue left Hitachi: its consumer-electronics IP was consolidated into and then transferred out to Hitachi Consumer Electronics → Hitachi Maxell → Maxell. Hitachi, Ltd. is not the current owner.

Assignment timeline

All six recorded events trace a pure corporate-reorganization chain (Hitachi family → Maxell family). No third-party LLC, no NPE, no security interest.

• 2000-03-08 (executed; date of filing) / recorded 2000-03-08 — Reel not retrieved (see caveat)

  • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: Takahiro Nakano; Ryuji Nishimura; Toshiro Kinugasa (inventors)
  • Assignee: Hitachi, Ltd.
  • Correspondent: not retrieved — flag: this is the field I could not read; it would be the prosecuting/signing agent of record for the 2000 filing.
  • Context: inventor-to-employer transfer at filing (standard obligation-to-assign).

• 2013-06-07 (executed) / recorded 2013-06-12 — Reel 030802/0610 (confirmed for sibling US 6,865,677; inferred, not independently verified, for '616)

  • Conveyance: Assignment
  • Assignor: Hitachi, Ltd.
  • Assignee: Hitachi Consumer Electronics Co., Ltd.
  • Correspondent: not retrieved.
  • Context: internal reorganization — Hitachi carving its consumer-electronics IP into the Hitachi Consumer Electronics entity.
  • ⚠️ Contradiction to flag: Maxell's own complaints in D. Del. 1:18-cv-00310 and E.D. Tex. 5:18-cv-00034 state "in 2009 Hitachi, Ltd. assigned intellectual property … to Hitachi Consumer Electronics." The recorded assignment date is 2013-06-07 (reel 030802/0610 per the sibling record). The 2009 date appears in the pleading narrative only; the assignment record says 2013. Prefer the recorded 2013 date per your source hierarchy.

• 2014-08-26 (executed) / recorded 2014-09-08 — Reel 033694/0745 (confirmed for sibling US 6,865,677; inferred for '616)

  • Conveyance: Assignment
  • Assignor: Hitachi Consumer Electronics Co., Ltd.
  • Assignee: Hitachi Maxell, Ltd.
  • Correspondent: not retrieved.
  • Context: internal reorganization — consolidation of the Hitachi CE IP into Hitachi Maxell to align IP with Maxell's licensing/R&D operation.

• 2018-01-25 — Reel not retrieved

  • Conveyance: Assignment ("ASSIGNMENT OF ASSIGNOR'S INTEREST")
  • Assignor: Hitachi Maxell, Ltd.
  • Assignee: Maxell, Ltd.
  • Correspondent: not retrieved.
  • Context: internal reorganization / name-and-entity alignment following the Oct-2017 holding-company restructuring. Reel/frame did not appear in the sources I could reach.

• 2021-11-29 — Reel not retrieved

  • Conveyance: Merger
  • Assignor: Maxell, Ltd. (the absorbed operating subsidiary, est. 2017-04-25)
  • Assignee: Maxell Holdings, Ltd. (surviving company)
  • Correspondent: not retrieved.
  • Context: recording of the 2021-10-01 absorption-type merger (confirmed by Maxell's own 2021-10-01 press release and the 2021-10-28 EDINET filing). Internal reorganization.

• 2021-12-03 — Reel not retrieved

  • Conveyance: Change of Name
  • Assignor: Maxell Holdings, Ltd.
  • Assignee: Maxell, Ltd.
  • Correspondent: not retrieved.
  • Context: change of name only — Maxell Holdings, Ltd. renamed itself "Maxell, Ltd." effective 2021-10-01 on conclusion of the merger. No change in beneficial ownership.

Correspondent-of-record finding (partial). I could not read the Assignment Center correspondent field for any entry. The strongest correspondent evidence I did surface is John R. Mattingly, Mattingly & Malur, PC (USPTO Reg. No. 30293), who signed the terminal disclaimer over '616 for Maxell's continuation application 14/264,243 ("ELECTRIC CAMERA," first inventor Takahiro Nakano, docket ASA-9606-06). Mattingly's firm is the recurring prosecution counsel for the Maxell/Hitachi camera family (e.g., Mattingly & Malur appears as attorney of record on Maxell/Hitachi applications). That is a prosecution correspondent, not a recording correspondent, so I do not treat it as satisfying the NPE "repeat recording correspondent" test — it is noted for completeness only. If you need the recording correspondents, they must be read directly from the Assignment Center reel images.

Timeline diagram

timeline
    title Ownership of US 6765616B1
    2000 : Filed by Hitachi Ltd
         : Inventors assign rights to Hitachi
    2004 : Patent issued
    2013 : Assigned to Hitachi Consumer Electronics
    2014 : Assigned to Hitachi Maxell Ltd
    2018 : Assigned to Maxell Ltd
         : Delaware infringement suits filed
    2019 : BlackBerry IPR settled
    2020 : Patent expired
    2021 : Merged into Maxell Holdings Ltd
         : Renamed Maxell Ltd

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. Every assignee in the chain is a named operating corporation: Hitachi, Ltd.; Hitachi Consumer Electronics Co., Ltd.; Hitachi Maxell, Ltd.; Maxell, Ltd.; Maxell Holdings, Ltd. No "IP/Licensing/Ventures" suffix, no Delaware/Texas single-member LLC, no registered-agent service address. Maxell, Ltd. self-describes (D. Del. 1:18-cv-00310 ¶1) as a manufacturer of storage media, batteries, optical components, etc., and its SEC-adjacent filings confirm ongoing product revenue.

2. Known asserter in the chain — NOT PRESENT as to the standard NPE list; PARTIAL as an active enforcer. The current owner Maxell, Ltd. does not appear on the Acacia / Marathon / IV / IPNav / Wi-LAN / Conversant / Vringo / Pendrell / Round Rock lists. However, Maxell is a high-volume patent plaintiff in its own right — it ran coordinated 2017–2018 campaigns (D. Del. 1:17-cv-01446; 1:18-cv-00216; 1:18-cv-00310; E.D. Tex. 5:18-cv-00136 and 5:18-cv-00034) and was hit by BlackBerry's IPR2019-00087 over this patent. This is operating-company assertion, not classic NPE activity — Maxell sells products and sues competitors (Olympus, ZTE, etc.).

3. Repeat correspondent across the chain — UNCLEAR. Assignment Center correspondent fields not retrieved. The only recurring name I can ground is Mattingly & Malur PC / John R. Mattingly (Reg. 30293) on the prosecution of the '616 continuation family (terminal disclaimer in App. 14/264,243 citing '616 as a prior patent). Insufficient to call it a recording-correspondent pattern.

4. Cascading transfers — NOT PRESENT. The transfers are spaced 2013 → 2014 → 2018 → 2021, driven by discrete corporate events (Hitachi carve-out, IP consolidation, 2017 holdco reorganization, 2021 merger). Not a <24-month chain of anonymous LLCs; no evidence the assignees share a correspondent address or common NPE principals.

5. Pre-litigation transfer — WEAK / EXPLAINED AWAY. The 2018-01-25 assignment to Maxell, Ltd. precedes the Delaware complaints (1:18-cv-00216, 1:18-cv-00310) by roughly one to two months, which superficially fits the "arranged to assert" tell. But the transferee is the same corporate group aligning ownership with its licensing entity — not a transfer to a new, unrelated asserter. I mark this unclear-but-benign, driven by the Jan-2018 record and the Feb/Mar-2018 filings; it does not indicate a spam-NPE setup.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 of any assignor or assignee appears in the record; the transfers are solvent internal reorganizations.

7. Privateering — NOT PRESENT. This was an intra-group IP consolidation, not an operating company feeding an NPE to sue its competitors. Hitachi did not convey to a third-party assertion vehicle; Maxell is a Hitachi-family successor that enforces in its own name.

8. Defensive aggregator — NOT PRESENT. The chain terminates at Maxell, Ltd., an operating manufacturer, not at RPX / AST / LOT / Unified / OIN. The patent has not been neutralized by defensive aggregation.

Verdict

Operating-company assertion.

The entire chain is an internal Hitachi→Maxell corporate reorganization — Hitachi, Ltd. (2000 reel, inventor assignment) → Hitachi Consumer Electronics (2013, reel 030802/0610 per sibling) → Hitachi Maxell (2014, reel 033694/0745 per sibling) → Maxell, Ltd. (2018) → Maxell Holdings (2021 merger) → Maxell, Ltd. (2021 name change). No shell LLC, no NPE-list assignee, no fire-sale, no defensive aggregator. The current owner Maxell, Ltd. is a product-shipping manufacturer that asserted the '616 patent against actual competitors (D. Del. 1:17-cv-01446 / 1:18-cv-00216 / 1:18-cv-00310; E.D. Tex. 5:18-cv-00136) and defended it in IPR2019-00087 (BlackBerry v. Maxell, settled). The only caveats are the un-retrieved Assignment Center correspondent fields and the un-retrieved reel/frame numbers for the original 2000, 2018, and 2021 recordings — the latter should be confirmed directly at https://assignmentcenter.uspto.gov/ before this is treated as final.

Verification link: https://assignmentcenter.uspto.gov/ — search 6765616.

I'll search for the prior art references cited in US 6,765,616 B1, focusing on the USPTO/Google Patents citation data.

I have partial citation data. Let me get the complete "Patent Citations" (References Cited) list from the patent record.

I have partial citation data but need the complete "(56) References Cited" list. Let me pull the front-page data from Justia and the patent PDF.

Scope and evidence limitations (read first)

Two things materially constrain this analysis, and I want to flag them rather than paper over them:

1. The authoritative full text you supplied does not include the front-page "(56) References Cited" block. The body text mentions only two prior-art documents by name — JP-A-9-270959 and JP-A-11-187306 — both discussed in the Background. Everything below about the cited references is therefore reconstructed from secondary databases (Justia's "Referenced Cited" tables for the continuation family, and a Google Patents citation-table snippet). I label each list by its source so you can see the provenance, and I flag where the two sources disagree.

2. I could not retrieve a clean USPTO (PatentCenter/PatFT) "References Cited" page for 6765616B1 in the available searches. Per your instruction I did not substitute results for similar numbers. The citation lists below should be treated as reconstructed, not as a certified USPTO front-page extract.

I am building on — not repeating — the earlier bibliography/claim-summary/litigation sections. One correction to the earlier section worth noting: the earlier section said Google Patents' "Patent Citations" was not separately retrieved; the searches now surfaced a partial citation-table snippet, which I report in Table C.

§102 framework and the critical dates for 6765616B1

Because the §102(b) bar date is 1999-03-08, only references published/patented before that date can be §102(b) art. References published in 2003–2008 (several appear in the family lists) cannot be §102(a)/(b)/(e) art against the '616 at all — they were almost certainly cited during prosecution of the later continuations, not the '616. This is the single most important filter to apply to the lists below.

Table A — U.S. patent documents in the family "Referenced Cited" list

Source: Justia "Referenced Cited" tables for US 9,544,517 B2 and US 8,339,493 B2 (family continuations). Titles were not returned by the database for these entries; descriptions below are inferred from the reference's field and are explicitly provisional.

Important §102 point: the post-1999-03-08 entries (5,986,698 onward) can only be §102(e) art, and only if their underlying US filing dates precede the '616's invention date. Several are simply the applicant's own family members cited in later continuations and are not prior art to the '616.

Table B — Foreign patent documents in the family "Referenced Cited" list

Source: same Justia tables (both 9,544,517 and 8,339,493).

Table C — Google Patents "Patent Citations" table for US6765616B1 (partial, as surfaced)

The Google Patents citation table for the '616 page returned this partial set (publication number / priority date / publication date / assignee / title). Note that this set differs from the Justia family list — see the discrepancy note below.

Cross-check: the appearance of JP H11-187306 A here independently corroborates that JP-A-11-187306 is a cited reference — it is also named in the specification body. This is the strongest corroboration in the whole dataset.

Discrepancy flagged: Tables A/B and Table C are different sets. Either (a) Google's table was truncated and mixed with the "Cited By" region of the page, or (b) the parent '616 and its continuations had different citations applied by different examiners. I cannot resolve this with the retrieved data; treat the union of A/B/C as the "possible citation set" and the intersection (JP H11-187306) as the only element confirmed by two independent sources.

§102 element-mapping — most relevant references

Caveat: §102 anticipation requires a single reference to disclose every element of the claim, including the claim-1/claim-13 hook (K1 ≠ K2 between two image types). I do not have the specification text of these references, so the mapping below is a provisional identification of which limitations each reference bears on, not a conclusion of anticipation. Where I say "potentially anticipates," I mean the reference is the type of art that a §102 challenge would rest on.

Tier 1 — most relevant (bear on the core mixing/culling + dual-mode concept)

• JP-A-9-270959 (pub. Oct 1997)

  • Citation: Japanese unexamined patent application, published Oct 1997 — Japanese Patent Office publication.
  • Date: §102(b) art (published well before 1999-03-08); also §102(a).
  • Description: Per the '616 specification itself, it "discloses an apparatus which mixes together or culls the pixel signals inside the image sensing device to reduce the number of signals to be read and therefore the read cycle."
  • Claims potentially anticipated: claims 1 and 13 (the "driver … to vertically mix or cull signal charges … of every K pixels" limitation), and claims 2 and 14 (mixing/culling inside the device to reduce output lines). The applicant distinguished it as still requiring field memory and a special sensor structure — which maps to the signal-processing-unit and display-unit elements of claims 1/13, so a §102 attack grounded on this reference would need to show the K1≠K2 and signal-processing elements too.

• JP-A-11-187306 (JP H11-187306 A, Sharp Corp., pub. 1999-07-09)

  • Citation: Japanese unexamined application, pub. 1999-07-09; corroborated by the Google Patents citation table (Table C).
  • Date: §102(b) (published 1999-07-09, before the 1999-03-08 bar? — careful: 1999-07-09 is after 1999-03-08, so it is §102(a), not §102(b); it qualifies only against the 2000-01-11/2000-03-08 invention date, not the one-year bar). Correction to intuition: this reference is §102(a) art.
  • Description: "Imaging device" — an image-pickup device; in the spec it is grouped with the prior art that generates television-conforming signals from large-pixel-count sensors.
  • Claims potentially anticipated: claims 1, 13 (image sensing device + driver + signal processing unit), and potentially claims 6, 18 if it shows mode-dependent readout.

• US 5,828,406 (Parulski et al., Eastman Kodak, issued 1998-10-27)

  • Date: §102(b) art.
  • Description: Electronic camera with selectable image-resolution/processing modes (title/field inference; not verified).
  • Claims potentially anticipated: claims 1 and 13 — specifically the K1 ≠ K2 conditional (two different images handled with different processing) and claim 6 / claim 18 (moving vs. still images with different factors). This is the reference I would rank first for an obviousness attack on the K1≠K2 hook, and the one most worth pulling in full.

• US 6,580,457 B1 (Eastman Kodak, pub. 2003-06-17; priority 1998-11-03) — "Digital camera incorporating high frame rate mode"

  • Date: publication is post-filing, so not §102(a)/(b) as a patent; it is a §102(e) reference only if its underlying US filing date precedes the '616 invention date (its 1998-11-03 priority suggests yes for a 1999 filing).
  • Claims potentially touched: claims 1, 6, 13, 18 — a digital camera with two operating modes (high-frame-rate vs. normal) is structurally the K1≠K2 concept.

Tier 2 — relevant to specific dependent limitations

• US 2002/0118291 A1 (Ishigami, pub. 2002-08-29; priority 1996-12-26) — §102(e) art only (published 2002; effective date is its 1996 provisional/priority filing). "Charge transfer device and method of driving the same, and solid-state imaging device and method of driving the same" — bears directly on claims 1, 2, 13, 14 (transfer/driving of a CCD and its phase-pulse timing).
• US 6,529,236 B1 (Fuji Photo Film, pub. 2003-03-04; priority 1996-09-20) — §102(e)-type only. "Digital camera for outputting digital image signals…" — bears on claims 1, 13 (signal processing unit / display unit) and claims 6, 18.
• EP 0840503 A2 (Olympus, pub. 1998-05-06) — §102(b) if its disclosure predates 1999-03-08 in relevant part. Field-inferred as image capture. Claims potentially touched: 1, 13.
• US 5,734,424 (Sasaki, 1998) and US 5,287,192 (Iizuka, 1994) — digital still/video camera art bearing on claims 1, 6, 9, 13, 18 (mode switching and still-image capture).
• US 5,847,758 (Iizuka, 1998) and US 5,986,698 (Nobuoka, §102(e) only) — image pickup art bearing on claims 1, 2, 13, 14.

Tier 3 — relevant to color-filter and architecture limitations

• US 4,054,915 (Sugihara, 1977), US 5,170,249 (Ohtsubo, 1992), US 5,187,569 (Tani, 1993), US 5,638,132 (Hokari, 1997) — solid-state color imaging / color-filter-array art bearing on the color-filter limitations of claims 1, 10, 11, 12, 13. These are the references a §102 challenge against the yellow/green/cyan or R/G/B vertical-stripe filter recitations would rest on.
• US 5,402,173 (Noguchi, 1995) and US 5,438,365 (Yamashita, 1995) — readout/transfer and signal-processing art bearing on claims 2, 3, 4, 5, 14, 15, 16, 17.

Most-relevant prior-art ranking (my assessment)

1. JP-A-9-270959 — the applicant's own closest-identified art for the mix/cull inside the sensor concept (claims 1/13 core; 2/14).
2. JP-A-11-187306 — the only citation corroborated by two independent sources; imaging-device art for the claim 1/13 combination.
3. US 5,828,406 (Parulski/Kodak) — the best candidate against the K1 ≠ K2 hook (claims 1, 6, 13, 18).
4. US 6,580,457 (Kodak, high-frame-rate digital camera) — secondary K1 ≠ K2 candidate (§102(e) timing must be checked).
5. US 5,734,424 / US 5,287,192 / US 5,847,758 — camera mode and pickup art (claims 1, 6, 9, 13, 18).

Explicit uncertainties (do not treat as settled)

1. The (56) block is not in the authoritative text you supplied. The two citation lists (Justia vs. Google) disagree; I cannot certify which references the examiner actually cited against the '616 as opposed to against its continuations.
2. I do not have the reference specifications, so the "which claim(s) it potentially anticipates" column is a field/title-based hypothesis, not a §102 element-by-element comparison. A real §102 analysis requires the full text of each reference and a limitation-by-limitation chart against claims 1 and 13.
3. The strongest limitation for any §102 attack is the K1 ≠ K2 conditional (claims 1, 6, 13, 18). None of the references I could characterize on this record is confirmed to disclose two different vertical mixing/culling factors for two different image types.
4. JP-A-11-187306's date (1999-07-09) is after the 1999-03-08 §102(b) bar, so it is §102(a) art, not §102(b) — a point that is easy to get wrong and that materially affects any invalidity theory.
5. The IPR2019-00087 petition (BlackBerry v. Maxell) would contain BlackBerry's actual §102/§103 grounds and exhibit list for the '616 — that is the single best primary source to consult for a real prior-art mapping, and I could not retrieve the petition body in these searches.

I'll ground this analysis in the actual prior-art references of record for the '616 patent before building the § 103 combinations.

Let me pin down the specific prior-art references cited on the face of the '616 patent.

Excellent — I've surfaced a PTAB filing that applies specific prior-art combinations against this family. Let me confirm the identities of those references.

Obviousness Analysis — US 6,765,616 B1 ("Electric camera") under 35 U.S.C. § 103

Governing law note: The '616 patent has a filing date of 2000‑03‑08 and a priority date of 2000‑01‑11, so pre‑AIA § 103(a) applies (as it did in the PTAB filings discussed below). References published more than one year before the filing date are § 102(b) statutory‑bar art; references published before the invention date but within one year are § 102(a) art. No § 103(c) common‑ownership exception is available, because all the references identified are third‑party art.

Note on the prior‑art corpus. The "Prior Art section" for this page consists of (i) the two Japanese publications the applicant itself cites and describes in the Background of the '616 specification, (ii) the U.S./JP references of record in the "Referenced Cited" list that was carried forward verbatim into the continuations of the same application (visible on the family member US 9,544,517 B2), and (iii) the prior‑art combinations actually applied against this family in the PTAB papers I surfaced. I have verified (i) and (iii) directly; for (ii) I have verified of record status but not each reference's full disclosure. Uncertainties are flagged at the end.

1. The prior art corpus (as actually grounded)

A. Admitted prior art in the '616 specification itself (weightiest art)

The applicant's characterization of these two references is an admission as to their disclosure, and admissions in the specification are prior art for § 103 purposes. I independently retrieved the JP‑A‑11‑187306 abstract (J‑GLOBAL), which corroborates and expands the applicant's summary: a CCD "having a pixel region larger than the pixel region corresponding to the standard television signal in both the horizontal and vertical directions" used as a shake‑correction pixel region; a normal mode that cuts out only the number of pixels corresponding to the standard television signal, vertically by operating the number of high‑speed transfer pulses with the timing generator and horizontally by frequency conversion; and a high‑resolution still mode that uses the maximum effective pixel region including the shake‑correction pixel region to output high‑resolution digital still data recorded in JPEG. It also discloses electronic zoom with inter‑field correlation motion detection. Source: https://jglobal.jst.go.jp/detail?JGLOBAL_ID=200903029109299328

B. References of record (carried into the family's "Referenced Cited")

From the US 9,544,517 B2 reference list (same specification): US 4,054,915 (Sugihara); US 5,170,249 (Ohtsubo); US 5,187,569 (Tani); US 5,287,192 (Iizuka); US 5,402,173 (Noguchi); US 5,438,365 (Yamashita); US 5,638,132 (Hokari); US 5,734,424 (Sasaki); US 5,828,406 (Parulski); US 5,847,758 (Iizuka); US 5,986,698 (Nobuoka); US 6,195,125 and US 6,519,000 (Udagawa); US 6,661,451 (Kijima); US 6,781,634 (Shiozaki); US 6,798,448 (Motono); US 6,906,746 (Hijishiri); US 6,970,191 (Sase); US 7,152,539 (Nishimura); US 7,352,391 (Hatakeyama); and JP 04‑323973, JP 9‑270959, JP 11‑004456 (Fuji Photo Film, "Static image input device," publ. 1999‑01‑06), JP 11‑187306, JP 11‑355665. Source: https://patents.justia.com/patent/[9544517](/patent/9544517)

C. The combinations actually asserted in the PTAB papers against this family

A petition/paper in the PTAB record (document 1510599) applies "Noboru '306" (= JP‑A‑11‑187306, matching the '306 suffix and its role as primary reference) as the primary reference, "Kijima '503" for a shutter button (element 46) and for reading out all pixels independently on trigger (Figs. 6 and 9), and "Fujimoto '435" for "color filters that pass the same colors, which are arranged vertically" (i.e., vertical stripes), specifically citing "col. 6, lines 40+" and Figs. 4 and 13 for yellow/green/white/cyan stripe filters used with a mosaic‑pattern solid‑state sensor. The petition's own claim‑mapping language tracks claims 1, 3, 6, 7, 8, 9 and 13–16 of this claim family almost element‑for‑element, including the gain‑adjust and image‑unstability limitations. Source: https://ptacts.uspto.gov/ptacts/public-informations/petitions/1510599/

That is important context: a third party has already built a § 103 case against this claim set out of these very references.

2. Level of ordinary skill

A POSITA here would be a person with a B.S. in electrical engineering (or equivalent) and roughly 2–5 years in consumer solid‑state imaging design, or the equivalent, familiar with: CCD vertical‑transfer clocking (three‑phase/six‑phase, transfer gates), NTSC/PAL field structure and the ~240 effective lines per field, high‑speed charge dump during vertical blanking, interlace phase relationships, and color‑stripe filter signal processing. This matters because every "gap" in the primary reference below is a routine engineering implementation under this skill level.

3. Claim structure to be tested

• Group I — core architecture: sensor N ≥ 3M vertically; in‑sensor vertical mix or cull every K pixels → M output lines, K ≤ ⌊N/M⌋; signal processor; display unit with the TV screen; and K₁ ≠ K₂ for two different image types (claim 1, and claim 13 with the "grid array / filters on grid pixels" framing).
• Group II — scheduling: read/discard the N − K·M excess rows during vertical blanking (claims 2, 14).
• Group III — aspect ratio & interlace: horizontal aspect‑ratio extraction + expansion over the full horizontal effective period (3, 15); shift mixed pixels each display cycle for interlace (4, 16); gravity‑center interpolation to exact 180° (5, 17).
• Group IV — modes: moving (K₁) vs. still (K₂), K₁ < K₂ (6, 18); no mixing, all pixels read independently on a shutter trigger (9).
• Group V — ancillary circuits: gain adjust tied to mixing count (7); shake detector shifting the extracted area (8).
• Group VI — color filters: Ye/G/Cy, Ye/W/Cy, R/G/B vertical stripes (10–12).

4. Ground 1 (primary): JP‑A‑11‑187306 in view of JP‑A‑9‑270959 → claims 1–5, 13–17

Why '306 alone does not anticipate. '306 teaches mode‑switching by extracting a sub‑area and reading all lines of that area, at a readout rate that can be two or more times the field period. It does not teach reducing the number of lines inside the sensor by vertical mixing/culling, and therefore does not reach the "mix or cull every K pixels" limitation, the K ≤ ⌊N/M⌋ relationship, or K₁ ≠ K₂ as a mixing‑factor difference.

Why the combination nonetheless renders Group I obvious.

Motivation to combine (KSR ¶¶ 5–6). The motivation is documented on the face of the art itself: '306's own approach requires reading all effective pixels over two or more field periods and then performing field‑memory time‑axis transformation. That is exactly the deficiency the '616 specification criticizes ("requires a large processing circuit, such as field memory … the image sensing device readout cycle is a plurality of times the field cycle, degrading the dynamic resolution"). '959 supplies the known technique — in‑sensor mixing/culling — that eliminates the field memory and preserves dynamic resolution. The combination is the fusion of two references addressed to the same field, addressing a problem the art had already articulated, with a reasonable expectation of success, yielding no more than the predictable sum of the elements. The claimed advantage ("without increasing circuitry such as field memory") is the motivation, and it is recited by the applicant.

Group III (claims 3, 4, 5 / 15, 16, 17). These are largely admissions plus routine practice:

• Aspect‑ratio extraction/expansion (3, 15): '306's frequency conversion block 7g performs horizontal frequency conversion, and its electronic zoom block 7b handles magnification — i.e., expanding a wider horizontal segment to fill the horizontal effective period is a disclosed function; electrical zoom with aspect correction was standard.
• Shifting which pixels are mixed each display cycle for interlace (4, 16): the '616 Background admits this is conventional — "the combinations of pixels to be cyclically mixed together are changed from one field to another to achieve the interlaced scanning," and that in NTSC cameras the area beyond the 480th pixel "is read out at high speed during the vertical blanking period."
• Gravity‑center interpolation to 180° (5, 17): once mixing an odd number of pixels is chosen, a fractional‑line phase error is inevitable; correcting it with two‑tap line interpolation (the '616 spec's own 19/20, 1/20 weighting) is the classic remedy and is taught in the of‑record line‑interpolation art (e.g., Iizuka, Sasaki) and contemplated by the non‑integer readout/interpolation architecture of '306.

Group II (claims 2, 14). '306 controls extraction position by "operating the number of high‑speed transfer pulses" — i.e., dumping excess lines fast, outside the effective period. The '616 Background confirms high‑speed vertical‑blanking dump of surplus rows is old. Obvious.

5. Ground 2: Ground 1 + Kijima (US 6,661,451; "Kijima '⑤03" in the petition) → claims 6, 9, 18 (and reinforces 1/13)

Kijima is used in the PTAB paper for exactly two limitations:

• A shutter/trigger device (element 46), and
• On trigger, no cyclic mixing — all pixels are read independently (Figs. 6 and 9).

That maps directly onto claim 9 and into the K₁ ≠ K₂ limitation of claims 1/13 and the moving‑vs‑‑still relationship of claims 6/18: the still‑image trigger state is simply a third mixing factor (K = 1, i.e., no mixing), which is also what '306's high‑resolution mode does (read all lines). With Kijima providing the trigger‑responsive mode switch and its all‑pixel readout, and '306 providing the still/moving mode pair, the claim‑6/18 ordering K₁ (moving) < K₂ (still) — 4‑pixel mixing for motion, 5‑pixel mixing for still monitoring in the disclosed embodiment — is a direct consequence of the disclosed teaching and not a patentable distinction.

Motivation: Kijima and '306 address the same design tension (a high‑pixel‑count sensor must both produce TV‑compatible images and support high‑resolution stills), and Kijima supplies the user‑triggered, no‑mixing, full‑pixel readout that a POSITA would want for the still capture. Nothing in either reference dissuades this.

6. Ground 3: Ground 1/2 + Fujimoto ("Fujimoto '435") [and, secondarily, US 4,054,915 Sugihara; US 5,828,406 Parulski] → color‑filter claims 10–12 and the claim‑13 "grid array / filters on grid pixels" framing

• Fujimoto '435 is cited in the PTAB paper for precisely the vertical‑stripe limitation — color filters "that pass the same colors … arranged vertically," taught as a "rather simple stripe filter used in combination with a mosaic pattern solid state image sensor," with Figs. 4 and 13 showing Ye/G/W/Cy‑type combinations. The petition expressly uses it to supply the stripe‑filter limitation that the '306+'⑤03 combination does not explicitly show.
• Claim 10 (Ye/G/Cy): admitted in the '616 spec as the conventional arrangement ("so‑called vertical stripes").
• Claim 11 (Ye/W/Cy): the '616 spec itself states the W‑for‑G substitution "a higher sensitivity can be obtained than the color filter arrangement of the previous embodiment" — an applicant‑supplied motivation, which is fatal under KSR (the rationale comes from the patent owner's own description).
• Claim 12 (R/G/B): the '616 spec again supplies the rationale "can directly produce the primary color signals … and can provide a camera with good color purity and good color S/N." Fujimoto '435's mosaic/stripe color‑filter teaching and the generic color‑filter art of record supply the structure.
• Claim 13's "grid‑arrayed pixels" wording adds no weight: any CCD/CMOS area sensor is a grid array, and color filters are necessarily formed on the pixels.

Motivation: vertical‑stripe filters are chosen exactly because they yield a fixed set of color samples per output line regardless of how many pixels are vertically combined — the '616 spec makes this point (so color processing works "no matter how many pixels are vertically combined"). The stripe‑filter choice and the mixing architecture are therefore mutually reinforcing, not independent inventions.

For the display unit limitation and the progressive‑scan/high‑resolution sensor backdrop, US 5,828,406 (Parulski, Eastman Kodak) — a reference of record — is a natural corroborating reference for a camera that reads a high‑resolution array differently for video‑rate output versus still capture. I flag that I have not independently re‑verified its full disclosure in this session (see § 9).

7. Ground 4: ancillary claims 7 and 8

Claim 7 (gain adjust keyed to the mixing count). When four pixels are summed the signal is 4× a single pixel; when three are summed it is 3×. Normalizing the analog front end to keep the downstream level constant is the predictable, necessary consequence of changing K — and the '616 spec proves it is merely conventional by describing the 4/3 gain ratio as a matter of course ("the gain of the gain adjust circuit 5 is increased to 4/3"). The PTAB paper likewise maps this limitation to gain‑adjusting circuits in Noboru '306 and Kijima '⑤03. Obvious; also arguably an inherent requirement of any working mixing camera.

Claim 8 (shake detector shifting the extracted area). This is squarely JP‑A‑11‑187306: angular‑velocity sensors (5a, 5b) → microcomputer (6) → shake‑correction value that vertically shifts the cut‑out by controlling the number of high‑speed transfer pulses and horizontally shifts it by frequency conversion. The PTAB paper cites '306 ¶0013 for exactly this. Obvious; in fact '306 alone discloses the substance of claim 8 (the only added wrinkle being that the size as well as the position of the extracted area can change, which is the necessary consequence of the variable‑K architecture).

8. Motivation‑to‑combine synthesis (why these combinations are proper)

1. Same field of endeavor, same problem, same solution type. All references address the mismatch between a high‑pixel‑count sensor and a fixed‑line TV standard.
2. The problem and the deficiency of the primary reference are recited by the applicant. The '616 Background states the field‑memory/time‑axis‑transformation drawback of '306‑type cameras in detail; that is a documented design incentive to adopt '959's in‑sensor mixing.
3. Known technique, predictable result. In‑sensor vertical mixing/culling to hit the field line count is a mechanical charge‑transfer operation with a predictable output; K = N/M arithmetic; gain normalization is arithmetic; interlacing by shifting mix groups is admitted conventional.
4. Two known options combined. '306 gives the two modes (TV‑rate with extraction; full‑resolution still); '959 and Kijima give the mixing/no‑mixing implementations. Selecting among a small set of known implementations to satisfy both modes is the essence of KSR.
5. No teaching away. Nothing in '959, Kijima, or Fujimoto '435 disparages in‑sensor mixing or full‑pixel still readout. The only tension — that '306 reads all lines slowly — is resolved, not aggravated, by the combination, and the '616's own spec concedes that faster in‑sensor reduction is a known remedy.

9. Where a validity challenge is weakest — defenses the patent owner will raise

These are the genuinely contestable points; a § 103 conclusion is not automatic:

1. The 3× floor (N ≥ 3M). '306's shake‑region sensor need not have 720+ vertical pixels for NTSC. The patent owner will argue that moving from the then‑conventional ~480‑line (2×) architecture to a 3×‑or‑more sensor, and only then mixing, was a deliberate and non‑obvious departure. Counter: the '616 spec concedes N may be "more than 1200," that N need not be divisible, and that the limit is driven by desired image‑stabilization area size — i.e., a result‑effective design variable (Aller). Expect this to be a factual fight about unexpected results.
2. The precise K ≤ ⌊N/M⌋ formulation and the "every K pixels" periodicity. A challenger must show the art reaches cyclic, uniform mixing/culling at a factor chosen to yield exactly M lines. '959 is described only generally; if '959's disclosure is thin on periodicity/selection of K, the owner will attack the combination as unsupported. This is the single most important reference to read in full.
3. "Display unit with the display screen of the television system" (claims 1, 13). If construed as a television monitor display screen rather than a camera viewfinder output, the owner may argue '306 discloses formats/outputs, not a display unit. Counter: viewfinder/LCD display units with TV‑timing output are ubiquitous (Parulski and the general camera art).
4. The "not cyclically mixed but read independently for all pixels" trigger claim (9) is very close to '306's high‑resolution mode; the owner's best argument is that '306 still reads in the frame‑transfer manner, not truly all pixels independently per frame. Kijima's Figs. 6/9 are the answer.
5. Claim 5/17's exact 180° correction may be argued as a specific solution not squarely taught; but line interpolation for interlace phase is standard practice, and the weighting (19/20‑1/20) is a trivial design choice once the 36° error is understood.
6. Contradiction to flag from the earlier section: the earlier-generated summary correctly noted the truncated text ended mid‑claim 18 and therefore did not characterize claims 19–20. Do not import claim language from US 9,544,517 (a continuation in the same family) into the '616 analysis — the '517 claims recite a zoom/magnification‑factor architecture with "K₁ larger than K₂" as magnification increases, which is not the '616 claim 6/18 relationship (there, moving K₁ < still K₂). Conflating the two would produce an incorrect element‑by‑element comparison.

10. Claim‑by‑claim disposition summary

11. Bottom line

Under pre‑AIA § 103(a) and KSR, claims 1–18 of US 6,765,616 B1 are, on this record, vulnerable to a well‑constructed obviousness challenge built on JP‑A‑11‑187306 (Noboru/Nakano '306) as the primary reference, in view of JP‑A‑9‑270959 for in‑sensor vertical mixing/culling, further in view of Kijima (US 6,661,451, "Kijima '⑤03") for the trigger‑responsive all‑pixel readout, and Fujimoto ("Fujimoto '435") for the vertical‑stripe color filters. The motivation is unusually well documented because the patent's own Background states the field‑memory/dynamic‑resolution deficiency of the primary reference, and the specification contains express admissions (high‑speed surplus‑line dump during vertical blanking; field‑to‑field mix shifting for interlace; Ye/G/Cy vertical stripes) plus applicant‑supplied rationales for the Ye/W/Cy and R/G/B variants. The dependent ancillary claims (7, 8) are the easiest to invalidate; the independent claims turn on how deferred the 3×‑floor and the K‑selection formulation are shown to be.

This conclusion is consistent with the real‑world posture: IPR2019‑00087 (BlackBerry v. Maxell, '616) settled, and separate PTAB papers in this family already advance these very reference combinations — see https://portal.unifiedpatents.com/ptab/case/IPR2019-00087 and the petition documents at https://ptacts.uspto.gov/ptacts/public-informations/petitions/1510599/.

12. Explicit uncertainties

1. Content of the of‑record U.S. references (Sugihara '915, Parulski '406, Udagawa '125/'000, Iizuka, Sasaki, Nobuoka, etc.) is asserted from their titles/assignees and general subject matter; I did not verify each specification's disclosure line‑by‑line in this session. Only JP‑A‑11‑187306 (abstract, retrieved), JP‑A‑9‑270959 (as the applicant characterizes it), and the references quoted in the PTAB petition paper have been directly grounded.
2. "Kijima '⑤03" is not confirmed to be US 6,661,451. The petition uses the short form "Kijima '503"; the family's of‑record Kijima reference is US 6,661,451. I could not confirm the correspondence from the retrieved snippet. If "Kijima '503" is a different Kijima patent, the trigger/all‑pixel‑readout mapping should be re‑verified against that document.
3. "Fujimoto '435" full number is unconfirmed. The petition quotes its disclosure (col. 6, lines 40+; Figs. 4, 13) but the retrieved snippet truncated the patent number.
4. "Noboru '306" = JP‑A‑11‑187306 is a strong inference from the '306 suffix, the publication date, and its role as primary reference; the petition's "Noboru" appears to be an inventor name, which the J‑GLOBAL record I retrieved did not list.
5. The proceeding that produced the petition paper (pxx 1510597/1510558/1510599) is not identified with certainty in the retrieved text; the language ("Claims 13‑16 are rejected under 35 U.S.C. 103(a) as being unpatentable over …") reads like a proposed rejection in a reexamination/reexam‑style filing. I have treated it as evidence of an actual applied combination rather than as a final adjudicated invalidity holding.
6. Litigation/appellate status of the '616 patent in 2026 remains as flagged in the earlier section — no confirmed 2026 CAFC docket tying this patent to an appeal; any such connection is unverified.