US 5646390

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

Let me run a couple more targeted searches on litigation and any 2026 appellate activity.

Patent Analysis Report — US 5,646,390

Scope note: I searched for the exact identifier 5646390 / US 5,646,390. Search results confirm this number corresponds to "Dataform readers and methods" (Wang et al.) and no other patent. Where sources conflict or where I could not verify, I say so explicitly below.

1. Bibliographic Data

Continuity (from the patent's own description): This patent is a continuation of application Ser. No. 08/332,592, filed Oct. 31, 1994, now U.S. Pat. No. 5,521,366, which is itself a continuation-in-part of application Ser. No. 08/280,489, filed Jul. 26, 1994, now U.S. Pat. No. 5,572,006. The 1994-07-26 priority date traces to US 5,572,006.

Assignment history (from the Google Patents record): Metanetics Corporation (1997) → Meta Holding Corporation (2000) → Telxon Corporation (merger, 2004) → Symbol Technologies, Inc. (2002/2004 chain, with security interests by The Bank of New York and JPMorgan Chase).

⚠ Minor source discrepancy: A third-party aggregator (patents.searchlight.law) renders the application number as "US 6224423" and lists an art unit of 254. The authoritative Google Patents record and the patent's own continuity text support application 08/622,442. I do not auto-correct the alternate figure; I flag it as a likely transcription artifact in that aggregator.

2. Abstract (as issued)

"Dataform readers and methods provide operation with a sequence of overlapping exposure periods for successive lines of sensor elements of an array. Also, the exposure periods for successive lines of sensor elements, or successive subsets of lines, can be independently determined. In operation of a dataform reader illumination of a target area is turned on before exposure of the first line of sensor elements, each line of sensor elements is exposed by reflected illumination in an exposure period which overlaps with the exposure period of one or more other lines, and illumination is then turned off. Using a CMOS construction, for example, the level of image signals read from one line of sensor elements is referred to a look-up table to determine an appropriate exposure period for a subsequent line or lines of elements of an array. This process is repeated to achieve exposure periods adjusted for localized exposure conditions for each successive line or subset of lines of sensor elements."

3. Plain-Language Overview of the Independent Claims

There are four independent claims: 1, 9, 13, and 15. (Claims 2–8 depend from 1; 10–12 from 9; 14 from 13; 16–20 from 15 and 17.)

Claim 1 — Dataform reader with overlapping, feedback-adjusted line exposures (apparatus)

A bar-code/dataform reader has a two-dimensional array of sensor elements arranged in lines. An array control assembly starts charge accumulation on a line when it receives an "exposure start" signal, and reads out that line's image signals (to an output point) on an "exposure stop" signal. An exposure control system watches the output point, uses the signal level read from already-exposed ("selected") lines to decide how long to expose a later line, and issues the start/stop signals. The key limitation: the exposure periods are sequenced so that one line's exposure window overlaps the next line's exposure window.

Claim 9 — Same concept, expressed around reference-potential reset (apparatus)

Substantially the same reader as claim 1, but the array control assembly is positively recited as initiating charge accumulation on a line that has first been set to a reference potential (i.e., grounded/reset before integration). Unlike claim 1, claim 9 does not require the overlapping-exposure-period limitation — it instead broadly requires that the exposure control system supply the start/stop signals to implement the line exposure periods. This makes claim 9 the broader of the two "control-loop" apparatus claims on that particular point.

Claim 13 — Complete reader: illuminator + focusing optics + exposure control + decoder (apparatus)

A full reader for reading a dataform in a target area at a distance. It adds to the array machinery: at least one exposure illuminator to light the target area, and a focusing device in front of the array that focuses reflected light when the distance is within a focus range. It then recites the array control assembly (start/stop signals, charge accumulation, readout to an output point), an exposure control system that sequences exposures so that one line's exposure overlaps a subsequent line's, and a processing unit that decodes the dataform from the image signals. Note: claim 13 requires the overlap feature but, unlike claim 1, does not require that exposure duration be derived from previously read image-signal levels.

Claim 15 — Method of reading with partially overlapping line exposures (method)

A method for a dataform reader having a sensor array, comprising:

• (a) causing illumination of a target area containing a dataform image to be reflected onto the array;
• (b) starting charge accumulation on a first line via an exposure start signal;
• (c) ending that first-line exposure period by reading its image signals on an exposure stop signal;
• (d) coupling the first line's image signals to an output point;
• (e) repeating (b)–(c) for successive lines, where each successive line's exposure is initiated during the preceding line's exposure period, producing partially overlapping exposure periods;
• (f) coupling successive lines' image signals to the output point in sequence; and
• (g) terminating the illumination.
• Dependent claims add: turning illuminators on/off (16); averaging a line's image signals to time a later line's exposure stop signal (17); using a look-up table for that timing (18); applying the averaged value to a line at least one line removed from the line sampled (19); and coupling image signals to memory and decoding the dataform (20).

4. Litigation / Docket Findings

• Family litigation flag: Google Patents records "Family has litigation," citing a U.S. district court case in the Texas Eastern District Court, case 2:05-cv-00456 (via Unified Patents' litigation dataset), and a "first worldwide family litigation" entry via Darts-ip (family 23298920). I did not retrieve a docket sheet, party list, or outcome for 2:05-cv-00456 in these searches, so I cannot state who sued whom or how it resolved.
• CAFC 2026 docket: I found no Federal Circuit docket, opinion, or order from 2026 involving US 5,646,390. Given the patent's status (expired 2014-07-26, anticipated expiration noted on the face of the record), 2026 appellate activity is unlikely. I cannot confirm any 2026 CAFC matter for this patent — treat the absence as "not found," not as proof none exists.
• Caution on a false positive: One search result appearing to place "5646390" inside a CAFC-adjacent document was in fact a prior-art citation table in an unrelated proceeding (a list of related patents cited as references). It is not a case about this patent.

5. Verification / Uncertainty Ledger

6. Technical Significance (brief)

The patent sits in the lineage of CMOS-based 2D dataform imagers developed by Metanetics (later swept up via Telxon into Symbol Technologies, and cited in later Symbol/Meta Holding imaging patents such as US 5,763,864 and US 5,717,195). Its distinguishing contribution relative to its parent US 5,521,366 is the rolling/overlapping line-exposure architecture — each sensor line integrates during a window that partially overlaps its neighbors — combined with a per-line (or per-line-subset) exposure feedback loop that samples the just-read line's average signal level and consults a look-up table to set the next exposure duration. This avoids the under/over-exposure gradient of simple sequential readout and the power cost of continuous-frame CCD operation. The CPC tag G06K 7/10752 (Exposure time control) captures this exposure-control thrust directly.

I'll research litigation involving US patent 5646390 using multiple sources.

Let me search targeted litigation databases and the specific case referenced on the patent page.

Let me research the specific case referenced and the assignee companies involved.

I'll start by verifying the ODP-sourced "no proceedings" baseline against public sources.

Initial searches surfaced mostly unrelated '390-family noise. Let me target the patent's actual litigation and any PTAB filings directly.

PTAB / AIA Trial Analysis — U.S. Patent No. 5,646,390

Patent: U.S. 5,646,390 — "Dataform readers and methods"
Inventors: Ynjiun P. Wang; Paul P. Ju
Original assignee: Metanetics Corp. Current assignees of record: Symbol Technologies LLC; Meta Holding Corp.
Priority: 1994-07-26 · Filed: 1996-03-25 · Issued: 1997-07-08 · Status: Expired – Lifetime (anticipated expiration 2014-07-26)

Method note: I ran the canonical structured "PTAB proceedings on file" block first (USPTO ODP: zero AIA trial proceedings), then ran four web searches to catch anything ODP hasn't indexed. The searches returned only unrelated patents that share the terminal digits "'390" (IPR2023-00475 on U.S. 7,116,390; IPR2013-00611 on U.S. 6,488,390; a POPR on U.S. 7,953,390) and unrelated Symbol/Metrologic litigation over different patents. None of those involve U.S. 5,646,390. I found no IPR, PGR, or CBM involving this patent.

Proceedings overview

Total AIA trial proceedings on file: 0 (zero IPRs, zero PGRs, zero CBMs — no institution decisions, no final written decisions, no settlements, no appeals). The patent is not "hardened" by surviving PTAB attack; it is simply untested at the Board and, more importantly, expired on 2014-07-26, which — combined with the six-year damages look-back of 35 U.S.C. § 286 — means a defendant receiving a demand letter today has an expiration/limitations defense so complete that the absence of PTAB history is largely academic. Any assertion grounded in §§ 271/284 is subject to a per se bar on recoverable pre-expiration damages.

Proceedings on file

None. There is nothing to report under the per-proceeding template (no proceeding number, no petitioner, no panel, no FWD, no appeal). I am flagging that plainly rather than filling the template with inference.

What I ruled out (so you don't chase it)

Why no AIA trial was ever realistically available

1. PGR — unavailable by statute. 35 U.S.C. § 321 applies to patents with an effective filing date on or after 2013-03-16. This patent's effective filing date is 1994-07-26.
2. CBM — doubly unavailable. CBM review under AIA § 18 was limited to patents claiming a method/apparatus for performing data processing or other operations used in the practice, administration, or management of a financial product or service; a CCD/CMOS dataform reader for reading bar codes does not qualify. Separately, the CBM program sunset on 2020-09-16.
3. IPR — legally available, practically foreclosed. An IPR could have been filed on any of claims 1–20 on a § 102/§ 103 ground based on patents or printed publications (§ 311(b) bars § 112 and public-use/on-sale art). But the patent expired 2014-07-26, removing the commercial incentive that drives IPR filings. The absence of IPRs is consistent with an expired, non-asserted patent, not with a validity challenge that failed.

Strategic summary

Claim status. All 20 claims of U.S. 5,646,390 stand UNTESTED — not canceled, not confirmed. No claim has been invalidated, narrowed, or affirmed by the PTAB or the Federal Circuit, so there is no estoppel-clean, claim-level roadmap from a prior FWD to work from. Note that the claim set in the record I have is truncated at claim 11, so I cannot give you claim-by-claim text for claims 12–20; pull the full issued claims from the USPTO PatentCenter or the Google Patents page before building any invalidity charts.

Estoppel landscape. There is no § 315(e)(2) estoppel to plan around, because no IPR was ever instituted. Equally, there is no petitioner-side prior art that has been "used up." A defendant today has the entire universe of § 102/§ 103 art and § 112 challenges available in district court — including art that could not be raised in an IPR at all. The relevant bars are temporal: § 286's six-year damages look-back plus the 2014-07-26 expiration.

Pattern signals. No serial petitioner, no serial patent owner, no defensive aggregator involvement. The patent is part of a 1994–1997 Metanetics/Dataform-reader family (including U.S. 5,572,006 and U.S. 5,521,366 as parents, and the single-chip CMOS imager application Ser. No. 08/258,428, filed 1994-06-10, referenced in the specification). The ownership chain is heavily consolidated: Metanetics → Meta Holding Corp. → Telxon → Symbol Technologies, with associated security interests with The Bank of New York and JPMorgan Chase. Current assignees of record include Symbol Technologies LLC and Meta Holding Corp. That is a large, sophisticated operating-company chain of title, not a classic troll structure — which further explains the zero-PTAB history.

One unverified lead you should chase before relying on anything. The Google Patents record for this patent (via the Darts-IP family map, family 23298920) flags litigation in the Eastern District of Texas, case 2:05-cv-00456. I could not confirm the parties, the asserted claims, or the outcome — my searches on that docket returned no reliable match, and I will not guess at party names. The Unified Patents litigation link is:
https://portal.unifiedpatents.com/litigation/Texas%20Eastern%20District%20Court/case/2%3A05-cv-00456
That is a district-court matter, not an AIA trial, and it should be pulled from PACER/CourtListener directly if a defendant is evaluating the patent's litigation pedigree.

Recommended next steps

If you are a defendant being asserted against today:

1. Lead with expiration, not invalidity. The patent expired 2014-07-26 (20-year term running from the earliest U.S. filing, 1994-07-26, per the record). There can be no infringement after that date. Any pre-expiration infringement more than six years before the complaint is barred by 35 U.S.C. § 286, and as of 2026 all pre-expiration conduct is outside the look-back window. This is a dispositive, low-cost motion.
2. Do not budget for an IPR. Filing an IPR on an expired patent is possible but economically pointless, and the Board's settled practice is to weigh expiration heavily. There is no FWD to cite, because none exists.
3. If you need a validity kill for license/portfolio-valuation purposes (not just this suit), your tools are an ex parte reexamination under 35 U.S.C. § 302 / 37 C.F.R. § 1.510 (any patent/printed-publication § 102/§ 103 ground, no filing deadline, no estoppel) or a district court § 112 / § 101 challenge — none of which PTAB AIA trials reach here.
4. Pull the full claim set and file history yourself before citing this patent. My source text truncates at claim 11 of 20, and the specification incorporates by reference application Ser. No. 08/258,428. If the demand letter cites specific claims, get exact claim text from https://patents.google.com/patent/US5646390/en or USPTO PatentCenter — do not rely on any secondary summary, including this one, for claim wording.
5. Treat the missing PTAB record as a data point, not a defense. The absence of IPRs is genuinely a signal — well-asserted, long-lived patents in the bar-code scanner space (as the Symbol/Metrologic ITC and E.D. Pa. matters show) did attract attacks — but here the more probable explanation is a patent that expired without ever being worth an IPR filing fee.

Constraints honored: I have not invented a proceeding number, petitioner, panel, or disposition. Where the record was truncated or unverifiable, I have said so explicitly. Authoritative source for the "zero proceedings" baseline is the structured USPTO ODP block in the task prompt; the patent text is from https://patents.google.com/patent/US5646390/en.

I'll research the assignment chain for US 5646390, cross-referencing the patent's recorded legal events with company histories and NPE databases.

Let me dig into the corporate history behind these assignments and check the USPTO assignment records.

I have the corporate backdrop. Now let me pin down the Metanetics / Meta Holding / Telxon relationship and try to recover reel/frame and correspondent data.

I hit the research step limit before I could pull the raw USPTO Assignment Center records (reel/frame and correspondent fields), so I'll be explicit below about what is documented vs. what I could not retrieve. Everything I state as a recorded event comes from the reassignment/legal-events table in the authoritative full text of US 5,646,390.

Assignment Chain Analysis — US 5,646,390 "Dataform readers and methods"

Scope note on sources. The authoritative patent text I was given includes the Google Patents "reassignment" event table (dates, participants, conveyance descriptions) but not reel/frame numbers, execution dates, or correspondents. Those fields live only in the USPTO Assignment Center record. I could not reach assignmentcenter.uspto.gov during this session, so reel/frame and correspondent are reported as "not retrieved" rather than guessed. Do not treat any absent field below as evidence that no record exists.

Inventors

Pattern note — inventor migration: At least one inventor (Wang) ended his career at Hand Held Products (Honeywell), a competitor of the eventual patent owner Symbol/Zebra, where he is named on the very generation of CMOS-imager patents that Honeywell asserted against Zebra/Symbol in 2022. I cannot date that move, so I cannot tie it to "departing within 12 months of filing" — it may be an ordinary multi-decade career arc rather than a fire-sale precursor. Flagging it as a cross-reference, not a finding.

Original assignee

Metanetics Corporation — named on the face of the patent; inventor rights assigned to it by the 1997-03-20 recorded assignment.

• Did it ship a product embodying the claims? Yes. Metanetics sold the IR-2000 "Hand Field Image Reader" and the IR-2000HD — documented in the exhibit/prior-art listing of US 8,397,992, which cites "Metanetics Corp., IR-2000 Hand Held Image Reader User's Guide, Release Date Nov. 1996" and "IR-2000HD ... User's Guide, Mar. 1998." That is exactly the hand-held 2-D dataform reader described in this patent.
• Primary line of business: hand-held image-based dataform readers / mobile data-capture hardware.
• Current status: no longer independent. Its IP passed to a holding/reorg vehicle (Meta Holding Corporation, 2000) and was subsequently consolidated into Telxon and then Symbol Technologies. The Google Patents page still lists Meta Holding Corp (alongside Symbol Technologies LLC) as a current assignee of record — see the note in the timeline below.

Assignment timeline

Dates are the event dates in the reassignment table. Because the table does not distinguish execution vs. recording, I present them as recorded. Reel/frame and correspondent were not retrievable from the sources available in this session.

1. 1997-03-20 (recorded) — Reel not retrieved

   • Conveyance: Assignment of assignors' interest
   • Assignor: Paul P. Ju; Ynjiun P. Wang (inventors)
   • Assignee: Metanetics Corporation
   • Correspondent: not retrieved
   • Context: Original inventor→company assignment (standard prosecution-time assignment; patent issued 1997-07-08).

2. 1999-03-30 (recorded) — Reel not retrieved

   • Conveyance: Security interest
   • Assignor: Telxon Corporation
   • Assignee (secured party): The Bank of New York
   • Correspondent: not retrieved
   • Context: Securitization — Telxon pledges its patent estate as collateral. (Telxon appearing as pledgor indicates the Metanetics portfolio was already inside the Telxon corporate family by 1999.)

3. 1999-09-09 (recorded) — Reel not retrieved

   • Conveyance: Release by secured party
   • Assignor: The Bank of New York (as agent)
   • Assignee/obligor released: Telxon Corporation
   • Correspondent: not retrieved
   • Context: Securitization unwind — release of the 1999 lien.

4. 2000-03-27 (recorded) — Reel not retrieved

   • Conveyance: Assignment of interest
   • Assignor: Metanetics Corporation
   • Assignee: Meta Holding Corporation
   • Correspondent: not retrieved
   • Context: Internal reorganization / transfer to holding entity — the operating business's IP moved into the "Meta Holding" vehicle.

5. 2002-04-08 (recorded) — Reel not retrieved

   • Conveyance: Assignment of interest
   • Assignor: Telxon Corporation
   • Assignee: Symbol Technologies, Inc.
   • Correspondent: not retrieved
   • Context: Acquisition cleanup — Symbol acquired Telxon in Nov. 2000 in a ~$465M stock-for-stock merger; this records the transfer of Telxon's patent estate to the acquirer.

6. 2004-12-22 (recorded) — Reel not retrieved

   • Conveyance: Merger
   • Assignor: Meta Holding Corporation
   • Assignee: Telxon Corporation
   • Correspondent: not retrieved
   • Context: Internal reorganization — Meta Holding merged into Telxon, which by then was a Symbol subsidiary; this is the step that pulled the Metanetics/Meta Holding line under Symbol ownership.

7. 2005-01-05 (recorded) — Reel not retrieved

   • Conveyance: Security interest
   • Assignor: Symbol Technologies, Inc.
   • Assignee (secured party): JPMorgan Chase Bank, N.A.
   • Correspondent: not retrieved
   • Context: Securitization — Symbol pledges its patent estate as collateral.

8. 2010-12-03 (recorded) — Reel not retrieved

   • Conveyance: Release by secured party
   • Assignor: JPMorgan Chase Bank, N.A.
   • Assignee/obligor released: Symbol Technologies, Inc.
   • Correspondent: not retrieved
   • Context: Securitization unwind — release of the 2005 lien. No subsequent ownership transfer is recorded.

Ownership-ordering caveat: events 5 and 6 appear chronologically inverted relative to the corporate logic (Symbol received Telxon's assignment in 2002, yet Meta Holding merged into Telxon only in 2004). The most defensible reading is that the 2002 Telxon→Symbol entry is a bulk assignment of Telxon's own estate, while the Metanetics/Meta Holding line reached Symbol via the 2004 merger into a now-Symbol-owned Telxon. I flag this as an interpretation, not a documented fact; the underlying instruments would need to be pulled from Assignment Center to confirm.

Timeline diagram

timeline
    title Ownership of US 5646390
    1994 : Application filed Jul 26
    1997 : Issued to Metanetics Corp
         : Inventors assign to Metanetics
    1999 : Telxon pledges security to BNY
         : BNY releases Telxon
    2000 : Metanetics assigns to Meta Holding
         : Symbol agrees to acquire Telxon
    2002 : Telxon assigns to Symbol Technologies
    2004 : Meta Holding merges into Telxon
    2005 : Symbol pledges security to JPMorgan
    2010 : JPMorgan releases Symbol
    2014 : Symbol becomes Zebra subsidiary

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only "Holdings"-style name in the chain, Meta Holding Corporation, is documented as merging into an operating company Telxon Corporation (2004-12-22 merger), which is the opposite of a licensing-shell endpoint. No assignee in the chain is a single-purpose licensing LLC, and the terminal owner (Symbol Technologies) is a large operating manufacturer.
2. Known asserter in the chain — not present. None of the assignees — Metanetics, Meta Holding, Telxon, Symbol Technologies — appears on the public NPE lists (Acacia, Marathon, IV, Wi-LAN/Conversant, Vringo, Pendrell, Round Rock, etc.). The one aggressive enforcer in the chain is Symbol, an operating company (see the 2005–2007 ITC action 337-TA-551 against Metrologic and its 2005 Delaware suits against Intermec, per the Symbol 10-K). That is operating-company assertion, not NPE conduct.
3. Repeat correspondent across the chain — unclear / not determinable. Correspondent fields were not retrievable in this session. This is precisely the field the task flags as the highest-value tell, so it should be pulled directly from Assignment Center before any conclusion is drawn. No correspondent name can be reported without fabrication.
4. Cascading transfers — not present (as an NPE pattern). There are four ownership events between 2000 and 2004, but they involve named, identifiable operating entities and documented corporate events (a $465M merger, a holding-company reorganization), not anonymous LLP-shuffling. The gaps exceed 24 months, and the participants are traceable, not shells.
5. Pre-litigation transfer — not present / unclear. The patent's family carries litigation (E.D. Tex. 2:05-cv-00456, per the patent page's "US case filed in Texas Eastern District Court" note). The nearest recorded ownership event, the 2002 Telxon→Symbol assignment, precedes a 2005 case by ~3 years — not within the 6-month window. I could not confirm from available sources which party filed 2:05-cv-00456 or whether US 5,646,390 itself was asserted; treat the "not present" call as provisional.
6. Bankruptcy fire-sale — not present / unclear. No bankruptcy docket for Metanetics, Meta Holding, Telxon, or Symbol surfaced. Note the 1999 security interest and its quick 1999 release, which can indicate lender pressure, but that is not a bankruptcy sale and the IP was later absorbed through the Symbol/Telxon merger at full value. The Chapter 11 case in this technology space belonged to Metrologic, an unrelated competitor.
7. Privateering — not present. No evidence the operating owner transferred the patent to an NPE to assert on its behalf. Symbol asserted its own patents in its own name.
8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Verdict

Operating-company assertion.

The chain runs from a genuine operating manufacturer (Metanetics, which shipped the IR-2000 and IR-2000HD readers), through documented corporate reorganizations and securitizations into the Telxon → Symbol Technologies merger (Symbol acquired Telxon for ~$465M in Nov. 2000), and terminates with Symbol Technologies LLC — now a wholly owned subsidiary of Zebra Technologies — as current assignee of record. Every assignee is a named operating entity; the only lienholders in the chain (Bank of New York, JPMorgan Chase) are secured lenders, not owners. No NPE signal is "present," so this is not an NPE chain.

Caveats: (a) the verdict describes the ownership chain, not proof that US 5,646,390 itself was asserted — I could not confirm the asserting party in E.D. Tex. 2:05-cv-00456; (b) reel/frame and correspondent were not retrieved and must be pulled from the Assignment Center before the chain is treated as fully verified; (c) the 2002/2004 ordering anomaly noted above should be resolved against the underlying instruments.

Verification links:

• USPTO Assignment Center: https://assignmentcenter.uspto.gov/
• USPTO assignment index (legacy): https://assignment.uspto.gov/patent/index.html
• Google Patents record (source of the event table): https://patents.google.com/patent/US5646390/en

I'll research the prior art cited on the face of US 5,646,390 and analyze each reference.

Let me search for the specific "References Cited" section on the patent's face.

Let me try to access the specific citation records for this patent.

Prior-Art Analysis for US 5,646,390 ("Dataform readers and methods," Wang & Ju)

⚠ Important scope/limitation statement — read first

The task asks me to work through "each patent citation for 5646390." I must be explicit about what I could and could not verify:

• I was NOT able to retrieve the examiner's front-page "References Cited" list (the numbered "U.S. Patent Documents / Foreign Patent Documents / Other Publications" block) for US 5,646,390 within my search budget. My searches returned the patent's title, dates, family, and forward-citation ("Citing Patents") data, but not the authoritative examiner citation block.
• Accordingly, I will not fabricate a citation list. Below I analyze (a) the references that US 5,646,390 itself cites inside its specification — which are verifiable directly from the authoritative patent text supplied — and (b) the family/continuity references. I flag each as to whether it can actually operate as § 102 prior art.
• One apparent "hit" for this number was a false positive and is flagged in § 4.

1. References cited within US 5,646,390 (verifiable from the specification)

These are the only references the '390 specification affirmatively cites. They are "prior art" in the sense of background/included-by-reference material, not necessarily examiner-applied art.

Specification location: Both are cited in the "Description of the Invention" as the mechanism for array construction/readout (Ref A) and for the decode subroutine (Ref B).

2. Family / continuity references (patent-to-patent relationships)

These are not "prior art" in the ordinary sense — they are the '390 patent's own ancestors.

⚠ Contradiction check against the earlier section: The prior "Patent summary" states the priority "traces to US 5,572,006." The family record supplied in the patent text lists 08/280,489 → patent/US5572006A/en, and a search result confirms US5572006A = "Automatic exposure single frame imaging systems." These are consistent — no contradiction between the spec text and the family record on the number itself.

3. § 102 analysis — which claims each reference could reach

Important doctrinal caveat (pre-AIA § 102, applicable to this 1994/1996 priority): references C and D are the same inventive entity (Wang & Ju) and commonly owned; they are therefore not available as § 102(a)/(e) anticipating art against '390. They matters for obviousness-type double patenting / § 103, not § 102 anticipation. Treat any "§ 102" framing of C/D with that caveat.

4. False positive — do NOT attribute this to '390

A search result surfaced an international search report for PCT/US 97/07033 (WO 1997/042021 A1) whose citation table listed DE 19 28 843 (Danzer), JP 05 017639 (Tokuyama Soda), WO 92/18173 (Exxon Chemical), EP 0 573 884 (Braun Melsungen), WO 93/23093 (Baxter), etc. Those references are all from the wound-dressing / polymer-chemistry field and have nothing to do with bar-code/dataform imaging. That ISR belongs to a different patent family; it is not an '390 citation list. (Source URL: https://patentimages.storage.googleapis.com/4b/2b/d8/daa7b404403d10/WO1997042021A1.pdf https://patentimages.storage.googleapis.com/4b/2b/d8/daa7b404403d10/WO1997042021A1.pdf)

5. What I could still confirm about the record (context, not prior art)

• The CPC tag G06K 7/10752 ("Exposure time control") is the classification most directly aligned with '390's exposure-control thrust — consistent with exposure control being the core contribution over parents C and D.
• Forward citations (patents that cite '390 — these are later art, NOT prior art to '390): e.g., US 5,691,773 (Wang et al., "Anti-hand-jittering dataform readers and methods," 1997-11-25), US 5,714,745, US 5,717,384, US 5,736,724, US 5,811,774, US 5,815,200, US 5,831,674, and the '390 family/grandchildren. (Sources: freepatentsonline.com/9578269.html; patents.justia.com/patent/7878403; patents.justia.com/patent/7594608) These are useful for later validity/landscape work but cannot anticipate '390.

6. Uncertainty ledger (this task)

Bottom line: Based on the verifiable text, the references that actually bear on US 5,646,390 are (A) Ser. No. 08/258,428 (sensor-array/readout — § 102(e)-eligible only if it granted; more likely § 103 support for the array dependent claims) and (B) US 5,304,787 (2-D code locating — a § 102(b) bar reaching only the decode/processing limitations). Neither appears to anticipate the independent claims' characterizing rolling/overlapping-exposure and per-line feedback-look-up-table features. To complete the assignment exactly as specified, the front-page examiner citation list (56/References Cited) would need to be pulled directly from the USPTO PatentCenter/Global Dossier record, which I could not access here.

I'll ground this obviousness analysis in the prior art actually associated with this page. Let me pull the cited references and check a few key prior-art patents.

Let me try to pin down the actual front-page references cited for this patent, since the page's "Prior Art" section includes cited-document data.

One more targeted attempt to surface verifiable third-party prior art in the line-exposure/feedback space before I write.

Obviousness Analysis — US 5,646,390 (“Dataform readers and methods”)

Under 35 U.S.C. § 103 (pre-AIA § 103(a)) — Application US 08/622,442 was filed 1996‑03‑25 with priority to 1994‑07‑26, so the pre‑AIA statute and pre‑AIA § 103(c) govern. (AIA first‑inventor‑to‑file provisions apply only to applications filed on or after 2013‑03‑16.)

1. Scope, sources, and evidentiary limits

I used the "Prior Art" material on the Google Patents record supplied in the full-text (source: https://patents.google.com/patent/US5646390/en), namely:

• Prior‑art keywords: sensor elements; exposure; line; array; image signals
• Prior‑art date: 1994‑07‑26
• References cited by number in the specification: U.S. Pat. No. 5,304,787 (“Locating 2‑D Bar Codes,” issued 1994‑04‑19, common assignee) and U.S. application Ser. No. 08/258,428, filed 1994‑06‑10 (common assignee, incorporated by reference).
• The specification’s own description of the prior art (continuous‑frame scanners, interlaced exposure, line‑by‑line shift‑register readout) — treated below as admissions under In re Nomiya / Riverwood Int’l v. R.A. Jones.

Caveat I must state plainly: neither the Google Patents record I was given nor my searches returned a front‑page "References Cited" (U.S. patent citation) list for US 5,646,390. The searches I ran returned mostly noise (unrelated PCT search reports, food‑science families, and a different Metanetics patent, US 5,521,366). I therefore do not name third‑party U.S. patent numbers as "cited references" unless I actually verified them, and I flag the two I did verify in §3. Where I rely on general knowledge of the art, I say so. I also note the supplied claim text is truncated mid‑claim 11, so my treatment of claims 12–20 relies on the previously generated summary (which I treat as authoritative here) plus the parallel claim set I verified in the parent, US 5,521,366.

2. Legal framework applied

Graham v. John Deere factors (scope/content of prior art; differences; PHOSITA level; secondary considerations), with KSR Int’l v. Teleflex, 550 U.S. 398 (2007) supplying the rationales: combining known elements according to known methods to yield predictable results; simple substitution of one known element for another; use of a known technique to improve a similar device in the same way; and "a finite number of identified, predictable solutions." MPEP 2143 supplies the checklist. Claim scope is measured by the broadest reasonable reading of each claim as issued.

3. The prior art actually available on this page

Documents expressly excluded from § 103 art: US 5,521,366 (parent, Ser. No. 08/332,592) and US 5,572,006 (Ser. No. 08/280,489), plus WO 96/03799 and WO 96/03708. These are in this patent’s own § 120 benefit chain / same Metanetics family and therefore do not predate the applicant’s invention as to the shared disclosure. I note that this removes the family’s look‑up‑table exposure teaching (e.g., US 5,572,006 cl. 30–31, 33) from the § 103 art — a point that cuts against the patent only indirectly, but it must be stated because an examiner-style combination cannot be built on those family members.

4. Element‑by‑element analysis

Claim 1 (overlapping, feedback‑adjusted line exposures)

Conclusion (claim 1): obvious. The only element not squarely admitted is staggered exposure starts, and the specification itself supplies the reason to adopt them: "because the exposure period required to appropriately expose a line of cells is longer than the time required to sample the cells … the controlled exposure period for the next line of sensor elements is initiated during the preceding exposure period." That is a throughput/pipelining observation, and KSR makes a predictable design‑driven optimization like this obvious. Note also dependent claim 6 (start signals at "predetermined uniform time intervals") — a uniform start cadence necessarily produces uniform, partially overlapping windows, confirming the overlap is an incidental consequence of the scheduling choice rather than a separate inventive insight.

Claim 9 (same, minus overlap; plus reference‑potential reset)

Claim 9 is the broadest claim in the patent: it drops the overlap limitation and simply recites reference‑potential (ground) reset — which the specification admits is the known structure of two‑dimensional photosensitive arrays (B). It therefore reads on admitted prior art plus per‑line exposure feedback (F, G). Conclusion: obvious — high confidence. This is the claim I would expect to fall first in any challenge.

Claim 13 (illuminator + focusing device + overlap exposure control + processing unit)

Every element except the overlap sequencing is conventional or in the record: LED exposure illuminators and LED/lens‑array illuminators (spec, FIG. 1A, and the family’s own earlier claim set), a central focusing lens with a depth‑of‑focus range (spec, FIG. 4 / distance 59, range 72), and a processing unit that locates and decodes 2‑D bar codes (D). Overlap sequencing falls for the reasons in claim 1. Conclusion: obvious — medium‑high confidence, the residual doubt being the breadth of "focusing device … within a focus range," which is plainly conventional optics.

Cross‑check that strengthens this: the parent, US 5,521,366, contains a near‑parallel claim (its cl. 17 — illuminator + focusing device + array control + exposure control providing start/stop "in a sequence causing the exposure period for one line … to overlap … a subsequent line" + processing unit). A child claim that merely re‑expresses the parent’s claim in slightly different words is not rendered non‑obvious by the rewrite.

Claim 15 (method: partially overlapping exposures) and dependents 16–20

• 15: each step is admitted art (A, B, C) plus staggered starts. Obvious — medium‑high.
• 16 (turn illumination on/off about the capture): conventional single‑frame capture; cf. F (processor‑controlled scan start/restart). Obvious.
• 17 (average a line’s image signals to time a later line’s exposure stop): summing/averaging pixel values across rows to derive an exposure parameter is conventional; the family’s earlier claim set does exactly this ("summing the pixel values represented by image signals from all sensors included in said two rows"). Obvious.
• 18 (look‑up table of exposure period vs. averaged level): routine controller implementation — a table of empirically derived values is the definition of a predictable design choice. Obvious (the added features of a general‑purpose controller do not confer patentability).
• 19 (apply the averaged level to a line at least one line removed from the sampled line): this is a scheduling choice, expressly discussed in the spec as an alternative (first‑line data used for the third line, or for a subset of three lines). Applying feedback with a fixed latency offset is the only way to overlap while still having feedback data available; G shows exactly this offset relationship (ni = nr − Te/tp). Obvious.
• 20 (memory + decode): D plus conventional memory. Obvious.

5. The three combinations I would run, and the motivation for each

Combination A — Admitted background (A+B+C) in view of the per‑line feedback exposure control of a CCD bar‑code reader (F).
Motivation: The Background itself frames the problem — continuous‑frame readers waste power; portable units are battery‑operated; non‑uniform illumination causes over/under‑exposure when one exposure is used for the whole image. A PHOSITA seeking a low‑power, single‑frame 2‑D reader would (i) keep the admitted X‑Y addressable array with reference‑potential reset and line‑by‑line shift‑register readout, and (ii) add the per‑line automatic exposure feedback that CCD scanner art already used (F), then (iii) stagger line exposure starts because the spec’s own admissions show the exposure window exceeds the readout time and because staggered windows are exactly what the admitted art already produced on the termination side (C). KSR rationale: known technique (per‑line exposure set from read‑out data) applied to a known device ready for improvement (the admitted line‑by‑line scanner), with predictable results (even exposure, no inter‑line gap in the output stream).

Combination B — Admitted interlaced/line‑by‑line array (B) in view of row‑offset (rolling) exposure scheduling in MOS imagers (G), optionally with feedback from F.
Motivation: G demonstrates that overlapping exposure windows are implemented by simply offsetting the row at which initialization occurs relative to the row being read, with the offset set by exposure time ÷ row period. Anyone who wants a per‑line exposure variable (claim 3’s premise) is driven to that offset scheme mechanically; there is no alternative that preserves a continuous output stream. KSR rationale: "obvious to try" a finite, identified set of scheduling options; predictable result.

Combination C — Combination A/B plus the located‑2‑D bar‑code decode of D.
Motivation: Decoding is the stated purpose of the reader (spec, step 117 and step 230), and the specification itself directs the reader to D for locating the dataform within the captured frame. Combining a capture front end with a known decoder that the applicant cites for the operative step is a textbook KSR combination ("familiar elements … according to known methods … predictable results"), and the reference to D removes any argument that the combination was unforeseeable.

6. Where the obviousness case is weak (and honest counterarguments)

1. The overlap element is the only genuine delta over the admitted art. The admitted art exposes "all or half the elements simultaneously"; it does not pre‑date interleaved per‑line start times. A patentee would argue that rolling exposure (as distinct from rolling read‑out) is a change in sensor architecture, not a mere administrative scheduling change. Rebuttal: the specification admits the read‑out side was already staggered and that the exposure window exceeds the read‑out time, making the start‑side staggering a result‑effective optimization; claim 6’s uniform start intervals concede the mechanism.
2. The strongest prior‑art hook (per‑line exposure from a look‑up table) sits in this patent’s own family (US 5,572,006’s look‑up‑table and averaging claims) and is therefore not available as § 103 art. An obviousness rejection built on those documents would fail on the § 120 benefit chain (and, for examination in 1996–97, common ownership would not have rescued the analysis either, since the pre‑1999 version of § 103(c) covered only § 102(f)/(g)). The case must therefore rest on A–C (admissions), F, and G, which is why I have grounded the motivation primarily in the specification’s own statements rather than in family documents.
3. I could not verify a "References Cited" list for this patent. If such a list exists and includes, e.g., a facsimile‑scanner per‑line AGC patent or a rolling‑shutter imager patent, the rejection would be considerably stronger than what I can support here. Treat my combinations as well‑motivated but not yet art‑complete.
4. Secondary considerations: I found no evidence of unexpected results, long‑felt need, or industry praise attributable to these claims as opposed to the broader family. The commercial chain (Metanetics → Meta Holding → Telxon → Symbol) and the Texas EDTX case 2:05‑cv‑00456 suggest commercial significance of the family, but I have no nexus evidence tying any success to the overlapping‑exposure limitation specifically, and I did not obtain the docket substance. Absent nexus, such evidence would carry little weight.

7. Bottom line

Overall: the claims of US 5,646,390 appear obvious under pre‑AIA § 103(a). The specification’s own Background does most of the work — it concedes the 2‑D sensor array, reference‑potential reset, shift‑register line readout, and even per‑line staggered exposure terminations — leaving only the staggered starts/partial overlap, which is a predictable throughput optimization that the specification itself explains (exposure time > read‑out time) and that the imager art already reduced to a fixed row offset. The proviso in §6.3 stands: I could not retrieve the patent’s front‑page cited‑references list, so this is a motivated prima facie case rather than a fully developed rejection.