US 4539585

I'll search for authoritative information on this patent number and any related litigation.

Let me run a few more targeted searches to check for litigation/docket activity and confirm the bibliographic record.

US Patent 4,539,585 — "Previewer"

⚠️ Threshold notes on scope and certainty

1. Identifier discipline. I searched literally for 4539585 / 4,539,585. Searches surfaced several coincidental string matches that are not this patent and should not be conflated:
   • Enanta Pharmaceuticals, Inc. v. Pfizer Inc., No. 2025-1427 (Fed. Cir. June 23, 2026) — this involves U.S. Patent No. 11,358,953 (the "'953 patent"), a coronavirus antiviral patent. That is a different patent number (11,358,953 ≠ 4,539,585).
   • An Argentine INPI trademark filing "Acta 4539585" (ARAN S.A.) and a Florida LLC FEI number "99-4539585" — unrelated.
2. No CAFC 2026 docket found for this patent. I found no Federal Circuit or district court docket in 2026 in which U.S. 4,539,585 is the patent-in-suit. This is consistent with its status (see below), but I state it as a negative search result, not a certification that none exists.
3. The patent has been expired since 2002-09-03 (per Google Patents' "anticipated expiration" entry), so it is not a plausible candidate for live 2026 infringement litigation.

---

Bibliographic summary

Sources: Google Patents US4539585A; FreePatentsOnline 4539585.

---

Abstract (verbatim from the patent)

"A previewer provides for the dynamic viewing of an article desired to be worn by the user in a complete series of poses without actually trying on the article. The image of the article is computer generated, and a large plurality of articles are available for previewing. The user pre-records, or creates in realtime, a series of poses which are time-framed and position orientation coded, which coded information is read by the computer and the computer generated article image is then re-imaged in accordance with the accessed coded information, and the re-imaged article image and the user video image are then video mixed, optically mixed, or superimposed in digital image storage memory, so that a user plus article composite image may be viewed by the user. The previewer is useful for previewing various articles or accessories, or combinations thereof, such as eyeglasses, hats, hairstyles, wigs, make-up, cosmetic surgical changes, clothing, furnishings, and automobile accessories."

---

Plain-language overview of the independent claims

Claim 1 — "A previewer" (apparatus).
A system that (a) uses a computer to generate a graphic image of an article (e.g., eyeglasses) rather than photographing a real one; (b) stores many video images of a subject plus a way to pick one specific frame; (c) attaches orientation reference means to the subject for each frame — the claim requires the indicia to encode x-y-z axes orientation and angular rotation about those axes; (d) has the computer read that orientation data and re-image the generated article to match the selected subject pose; (e) composites the re-imaged article with the subject image; and (f) displays the composite. In short: computer-render the product, track the subject's actual head/body pose, and merge the two.

Claim 13 — "A fashion previewer" (apparatus, garment-based).
A variant where the orientation reference is a garment bearing many coded indicia (each with a garment orientation reference and, in the disclosure, an equilateral triangle plus a binary reference number). A video camera photographs the coded garment as worn; a digital image processor converts the codes into digital instructions; a computer generates wearing-apparel image information per those instructions; and the system displays the re-imaged apparel as if worn by the subject. Again, the orientation indicia must provide x-y-z axes orientation and angular rotation about those axes. This is the "virtual dressing room" claim.

Claim 20 — "A previewer" (apparatus, real-time/transmission-oriented).
A generally broader/cleaner statement of the core idea: computer means to generate graphic image information of an article; means to transmit a video image of a subject carrying orientation reference information; means to process that video; computer means to input the orientation reference information and re-image the article accordingly; and means to display the subject image with the re-imaged article as a composite. The orientation reference information must again comprise x-y-z axes orientation and angular rotation indicia. Claim 20 is the ancestor for its dependents 21–23, which add (21) means to alter the article image beyond its originally defined parameters, (22) a joystick for inputting subject orientation, and (23) a keyboard for altering the article image.

Note on the dependent set: the remaining claims are dependent and add elements such as frame identification (claims 2–3, incl. binary digital frame IDs), recording/playback of subject video (4–5), a video mixer for compositing (6), article selection (7), joystick control (8–9, including a joystick shaped like a head), "fashion previewer" framing for human subject + plural articles (10), a single geometric shape as the indicia (11), indicia on one part of the subject (12), continuous/realtime re-imaging with subject movement (14), and apparel selection (16), indicia identification (17), binary reference number identification (18), and a form-fitting garment (19). Also note an apparent internal dependency oddity: claim 15 recites "The fashion previewer of claim 12" even though it recites garment orientation-reference subject matter; I report the claim text as printed rather than "correcting" it.

---

Why this patent surfaces in modern records (it is cited as prior art, not litigated)

• PTAB / IPR: US 4,539,585 (Spackova) was relied on as a prior-art reference (Ex. 1003, "Spackova") in IPR2013-00432, a challenge to U.S. Patent No. 6,624,843 (Lennon Image Technologies), typically combined with U.S. Patent No. 4,731,743 to Blancato. See PTAB petition documents.
• District court: in Luxottica Retail North America, Inc. v. Lennon Image Technologies LLC, "U.S. Patent No. 4,539,585 to Spackova et al." appears as petitioner's Ex. 1003 — i.e., as a reference, not the patent-in-suit. See Patexia docket.
• It is also repeatedly cited as an "A" (background) reference in later search reports (e.g., WO 01/46911, WO 97/43020, WO 03/005306, EP 2 113 881).

---

Stated uncertainty

• No independent corporate assignee: the record lists the assignee as "Individual" and shows no patent family members ("NONE") — I did not verify chain of title through USPTO Assignment records, so I cannot rule out an unrecorded or later-recorded assignment.
• CAFC 2026: I could not locate any 2026 Federal Circuit appeal naming US 4,539,585. Given the 2002 expiration, I would treat any such reference as highly unlikely, but I cannot affirm a definitive negative from the search tools available to me.
• Date of review: searches were run against the current date. Note that your prompt states April 26, 2026 while my working date is September 27, 2026; this does not affect the bibliographic data above, which is drawn from the patent document itself.

I'll search for litigation involving US Patent 4,539,585 specifically.

Let me search patent litigation databases specifically for this patent number.

Summary of findings: US Patent 4,539,585 — no known litigation

Bottom line: I could not identify any litigation (district court, ITC, or CAFC appeal) involving US Patent 4,539,585 as an asserted patent. The searches returned only the patent's own bibliographic record, prior-art citations, and unrelated hits from other patents/patents-databases.

Patent identity (for confirmation) — as retrieved, not corrected

Searches performed and what they returned

1. Unified Patents portal — https://portal.unifiedpatents.com/patents/patent/US-4539585-A. The page returned only patent metadata, "Referenced By (403)," and non-patent literature; no litigation list was returned for this patent number.
2. Google Patents — https://patents.google.com/patent/US4539585A/en. The record contains "Cited By (195/391)" and "Citations (4)" sections (i.e., patent citations), but no litigation / "Litigation" section and no "Litigation Status" entries.
3. Docket Alarm and general web searches for "4539585", "4,539,585", "Previewer" patent lawsuit — no district-court or CAFC docket asserting the '585 patent.
4. PTAB / IPR searches — no IPR, PGR, or CBM proceeding identified on the '585 patent.
5. CAFC searches — no Federal Circuit appeal identified naming the '585 patent.

Relevant "near misses" — expressly NOT litigation on the '585 patent

• Lennon Image Technologies, LLC v. Ulta Beauty, Inc., No. 2:20‑cv‑00249 (E.D. Tex., exhibit filed Aug. 4, 2020), https://www.docketalarm.com/cases/Texas_Eastern_District_Court/2--20-cv-00249/Lennon_Image_Technologies_LLC_v._Ulta_Beauty_Inc/1/1/. The '585 patent (spelled "spaekova et al.") appears as a reference in an exhibit/claim chart, not as an asserted patent of record. The patents-in-suit there appear to be different (Lennon-family) patents. This is a prior-art reference cite, not an infringement assertion of '585.
• WO2006134037A1, WO2001046911A1, WO2003005306A1 (and others) cite US 4,539,585 A as prior art in their International Search Reports (e.g., "X | US 4 539 585 A (SPACKOVA ET AL) 3 September 1985"). These are prosecution citations, not litigation.

Why an absence of litigation is plausible here

• The patent expired (for failure to pay maintenance fees) with an anticipated expiration of 2002‑09‑03, and the term (17 years from grant) ran out on 2002‑09‑03. Any enforcement suit would have had to be filed before expiration.
• It was never assigned to a corporate/operating entity (original and current assignee = "Individual"; family = "NONE"), reducing the likelihood of a portfolio-enforcement campaign.
• No corresponding family members exist (per EP/WO search reports showing "US 4539585 A … NONE"), so there is no parallel foreign litigation to point to.

Caveats on completeness

• Litigation databases (Unified Patents, Docket Alarm, PACER-derived) have incomplete historical coverage of the 1985–1995 period and of low-value/individual-owner suits. I cannot exclude the theoretical possibility of an early, unreported 1980s–1990s infringement action that never appeared in the indexed sources. I found no evidence of one.
• Google Patents and the Unified Patents page for this patent returned no "lawsuits"/litigation data fields at all, which is consistent with (but not conclusive proof of) zero known litigation.
• I did not find any AIA post-grant proceedings (IPR/PGR/CBM) or ITC § 337 investigations involving this patent number.

Conclusion: Based on the searches run, there is no known litigation involving US Patent 4,539,585. If you need a definitive negative, the next step would be a paid, full-text PACER/LEXIS/Westlaw docket search across all districts and the Federal Circuit from 1985 to 2002, plus a manual review of the assignment/chain-of-title record for Daniela S. Spackova and Richard M. Chen — sources I cannot fully query from the tools used here.

I'll do a quick check for any PTAB activity that the ODP ingest might have missed.

Proceedings overview

Zero AIA trial proceedings are on file for US 4,539,585 as of the USPTO ODP ingest date. The structured source block for this patent returns no IPR, PGR, or CBM records — no instituted trials, no denied petitions, no settlements, no claim cancellations, no Federal Circuit appeals. Breakdown: 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution-denied. My independent web checks (PTAB E2E/PTAB-ACTS public document hits, CourtListener, Fed. Cir. docket-adjacent sources) surfaced no petition or decision naming the '585 patent either, so I have nothing to flag as recently-filed-but-unindexed.

The bottom line for a defendant is not "the patent survived two IPRs and is hardened" — it is the opposite and more absolute: the patent's terminal date has passed and it was never tested at the PTAB at all. US4539585A carries legal status "Expired - Fee Related", with an anticipated expiration of 2002-09-03 (17 years from the 1985-09-03 grant, under the pre-URAA term then in effect). The AIA trial regime did not exist until 2012-09-16. Any demand letter asserting this patent today is asserting an expired patent whose last possible act of infringement occurred more than two decades ago.

No proceedings to report

There is no IPR####-#####, PGR####-#####, or CBM####-##### number to list, because no such proceeding was ever filed. Per the operating constraints, I will not invent proceeding numbers, panels, FWD dispositions, or appeal dockets to populate the per-proceeding template. The template fields (type, filed date, status, judge panel, grounds, institution decision, FWD, settlement, appeal, defensive value) have no data behind them for this patent.

Verification note / caveat: I confirmed the absence of AIA trial activity only. Ex parte reexamination and reissue are not AIA trials and are not part of the structured ODP block; I did not find evidence of either for the '585 patent, but I did not run an exhaustive CRU (Central Reexamination Unit) file-history check, so I state that as "no evidence found," not "confirmed none." I also found no district court or ITC litigation asserting the '585 patent. If you need certainty on the litigation piece, a PACER/Lex Machina docket sweep for 4,539,585 is the appropriate next step.

---

Strategic summary

Claim status of the 23 claims of US 4,539,585: all UNTESTED, none canceled, none sustained, none narrowed. There is no PTAB claim-level adjudication to report — independent claims 1, 13, and 20 and all twenty dependent claims stand exactly as they issued on 1985-09-03, untouched by any post-grant process. (Two drafting artifacts worth knowing if you ever have to read the claims: claim 15 recites "The fashion previewer of claim 12," but claims 12 is a dependent apparatus claim and 13 is the "fashion previewer" claim — apparently a typographical dependency error; and claim 6 depends on claim 5, while the "virtual dressing room" descendants in this space typically map to the composite-image claims 1 and 20.) The surviving claim set is listed in the authoritative text as claims 1–23, with claim 1 (computer-generated article image + stored subject frames + subject orientation reference providing x-y-z axes orientation and angular rotation orientation + re-imaging + composite display) as the broadest teaching.

Estoppel landscape: empty. Because no IPR, PGR, or CBM was ever instituted, 35 U.S.C. § 315(e)(2) estoppel never attached to anyone. There is no petitioner, no privity chain, and no "grounds raised or reasonably could have been raised" that are now foreclosed in district court or before the ITC. Practically, that means every prior-art ground, every § 102/§ 103 combination, and every § 112 theory remains fully available in any forum that could conceivably hear a validity challenge. The structured record lists four references cited by the examiner — US3510210A (Xerox, 1970-05-05), US3621214A (Romney, 1971-11-16), CA1020276A (Ryan, 1977-11-01), and EP0024862A2 (Taylor, 1981-03-11) — and none of them has been litigated or adjudicated. Conversely, because the patent expired on 2002-09-03 and the § 286 six-year damages bar runs from filing, there is no live monetary exposure to defend against; the estoppel analysis is largely academic.

Pattern signals: none of the usual troll/IPR hallmarks. No serial petitioner, because there is no petitioner. No patent-owner PTAB appeals, because there is no Board decision to appeal — and the record identifies no assignee beyond Individual (inventors Daniela S. Spackova and Richard M. Chen), so there is no corporate patent owner driving enforcement. No defensive aggregator (Unified Patents, RPX, etc.) in the chain as a challenger; the patent does appear in Unified's public patent portal and on Google Patents, but as a citation/prior-art node, not as the subject of a filed challenge. The real modern significance of this patent is its role as prior art, not as an asserted right: Google Patents shows it in the "Cited By" chains of ~195 first-listed publications (with an expanded second tally of ~391 — I note the discrepancy in the source rendering rather than picking one), and the EPO cited US 4,539,585 A as X-category art (relevant alone) against a later virtual try-on application, in a search report that also X-cited Longino's US 2005/0035980 and WO 00/16683 (Virtual Visual Devices). That is the posture a practitioner should expect: the '585 disclosure is the sword, not the shield.

---

Recommended next steps

• If you are a defendant and a demand letter or complaint cites US 4,539,585: you do not need an IPR. Lead with expiration. The authoritative record states legal status "Expired - Fee Related" with anticipated expiration 2002-09-03. An expired patent cannot be infringed, and any damages theory is independently barred by 35 U.S.C. § 286 because no act of infringement within six years of suit is possible on a patent that terminated in 2002. There is no FWD to link to, because none exists — say that plainly in your response rather than implying otherwise.
• On forum timing: § 315(b)'s one-year IPR clock is triggered by service of a complaint alleging infringement. With no complaint (none found) and no live term, that clock is irrelevant, and there is no statutory trial deadline, institution-decision deadline, or FWD due date to track — no AIA trial is pending.
• The absence is itself the signal, and it cuts your way. As a general matter well-asserted patents eventually attract IPRs; this one attracted none across the entire life of the AIA regime, which is consistent with a patent that was already expired before the AIA trial regime began and whose disclosed subject matter had migrated into the prior-art base. Treat the '585 patent as a freedom-to-operate / invalidity reference against other people's patents, not as a threat.
• Belt-and-suspenders checks: (1) PACER/Lex Machina docket sweep for 4,539,585 to confirm zero district court/ITC assertions; (2) a CRU reexamination/reissue file-history request via USPTO Patent Center to confirm no ex parte reexam certificate issued (this would be a narrowing of claims and could matter only if someone is claiming through it). Neither check changes the expiration analysis, but both close out the "no evidence found" caveats above.
• Do not draft around this patent. If a client is building virtual try-on / eyewear-preview technology, the productive use of the '585 patent is as anticipated prior art and as a § 102/§ 103 citation against newer claims — consistent with how the EPO used it against EP 2 113 881 A1 (search report at https://patentimages.storage.googleapis.com/0d/75/2d/a101bbb1a15dce/EP2113881A1.pdf).

Sources: authoritative full text — https://patents.google.com/patent/[US4539585A](/patent/US4539585A)/en (status "Expired - Fee Related"; anticipated expiration 2002-09-03; 23 claims; four examiner-cited references). PTAB proceedings list — USPTO Open Data Portal structured block supplied with this prompt (returns zero AIA trial records for this patent). No PTAB decision, no PTAB E2E paper, and no CourtListener/Federal Circuit opinion naming US 4,539,585 was located.

Preliminary data gathering

Let me verify the record before writing anything.

Let me try a couple more targeted probes on the assignment databases.

Assignment Record Reconstruction — US 4,539,585 ("Previewer")

Methodology note / limitation up front. I was unable to load the USPTO Assignment Center result page (https://assignmentcenter.uspto.gov/) directly in this session — that interface requires an interactive query and does not expose a crawlable per-patent result page. What follows is reconstructed from the corroborating records I could reach: Google Patents legal-events and assignee fields, the EPO/Espacenet bibliographic record and its patent-family annex, FreePatentsOnline, and the PTAB/ITC filings that cite the patent. Where the primary source is silent, I say so rather than fill the gap. The two search-engine result sets for the assignment databases returned only unrelated patents (Cree, Blue Spike, Genentech, University of Alberta) — no hit for 4,539,585.

---

Inventors

Unusual patterns: Yes, one — but not the one you asked about. Because there was no original corporate assignee, there is no "inventors departing the assignee" signal to measure. The pattern here is the absence of any employer at all: a two-inventor, individually-owned, privately-prosecuted filing in a capital-intensive field (real-time digital image processing, per the specification's references to a Vicom 16-bit image processor and a CAT-100 image processor). That is consistent with independent inventors or a startup that never recorded a corporate assignment — not with a corporate portfolio.

---

Original assignee

None recorded. Google Patents lists both Original Assignee: Individual and Current Assignee: Individual. Every bibliographic source I checked (Espacenet, FreePatentsOnline, Unified Patents portal) shows the inventors as the applicants, with no corporate applicant of record.

• Product embodying the claims: No evidence of a shipped commercial product. The specification describes a conceptual system assembled from then-available components (Vicom System Inc. and Digital Graphic Systems image processors) rather than a manufactured device.
• Primary line of business: Not applicable — no entity.
• Current status: Not applicable. The patent itself is Expired – Fee Related, with an anticipated expiration of 2002-09-03 (end of the 17-year term from the 1985-09-03 grant).
• Patent family: The EPO search-report annexes for two later applications (WO 01/46911 and WO 91/07727) state "US 4539585 A — 03-09-1985 — NONE" for patent family members. There were no foreign counterparts filed. That is a further indicator of an individually-financed filing rather than a corporate portfolio strategy.

---

Assignment timeline

The USPTO Assignment Center has no recorded assignments for US 4,539,585. There is no reel/frame to cite — not because the record is inaccessible, but because no post-filing conveyance (assignment, security agreement, merger, change of name, license, release or correction) has been indexed for this patent number in any of the sources I reached.

Consequently there is no correspondent of record to capture, and no assignee chain to diagram. This null result is itself the finding: the patent apparently remained in the hands of the two individual inventors (or their estates) from the 1981-07-10 filing through expiration in 2002.

Two events that do appear on the legal-events record, and which are not assignments:

• 1981-07-10 — Application filed by Individual (US 06/282,055).
• 1985-09-03 — Patent granted; publication of US 4,539,585 A.
• 2002-09-03 — Anticipated expiration.
• Current status — Expired – Fee Related.

(Per your instructions, the assignment enumeration stops here — but I am continuing to the remaining requested sections because the "Insufficient data" verdict option exists precisely for this scenario.)

---

Timeline diagram

timeline
    title Ownership of US 4539585
    1981 : Application filed by individual inventors
         : No corporate assignee named
    1985 : Patent granted to inventors as individuals
    2002 : Patent expired for failure to pay maintenance fees
         : No assignments ever recorded

---

NPE / troll-pattern signals

One counter-signal worth recording, in the opposite direction. Far from a troll pattern, this patent is a heavily-cited foundational reference in the virtual try-on art: 195+ forward citations per Google Patents, including Asahi Glass (US 4,730,260), Hoya (US 4,845,641), L'Oréal (US 4,670,781), Levi Strauss, eBay, and the My Virtual Model / Jarbridge families. Its holders never monetized it against any of them. A patent-assertion vehicle would not have left that citation stream — and the resulting claim-charting value — unexploited.

---

Verdict

Insufficient data — specifically, a null result: no assignments of record, not an incomplete record.

Justification: (1) no reel/frame entry exists for US 4,539,585 in any indexing source I reached, and both Google Patents assignee fields read "Individual" for original and current assignee; (2) the patent issued 1985-09-03 to two unaffiliated individual inventors, has no foreign family members per the EPO annexes, and expired 2002-09-03 for fee non-payment; (3) there is therefore no shell LLC, no repeat NPE correspondent, no cascading transfer, no pre-litigation transfer, and no defensive-aggregator termination to evaluate. The absence of any recorded transfer over a 21-year term is the substantive finding: this patent was never part of an NPE chain. The only assertion activity touching it runs the other way — it served as prior art against Lennon's US 6,624,843 in IPR2013-00432.

Verify at: USPTO Assignment Center (search "4539585") · Google Patents legal events · Espacenet bibliographic record

Caveat on confidence: I could not execute the Assignment Center query myself in this session. If you run it and it returns a reel/frame that the public indexes have missed, the sections above should be re-run — but the convergence of Google Patents, Espacenet and FreePatentsOnline all showing no assignee entity makes a missing assignment chain unlikely.

I'll research this systematically. Let me start by pulling the USPTO record for the patent and the four patent citations listed on its face, plus the references discussed in the specification.

Prior-Art Analysis — US Patent 4,539,585 ("Previewer")

1. Confirmation of the target patent (searched literally as 4539585)

I searched for US 4,539,585 only (no similar/neighboring numbers). The record was returned consistently across the USPTO-derived mirrors (Google Patents copy of the USPTO front page, FreePatentsOnline, uspto.report). Bibliographic data verbatim:

Note: the "Cited By (195 / 391)" documents on the Google/Unified pages are forward citations that postdate the 1981-07-10 priority date and therefore are not §102 prior art against the '585 patent. The only true prior-art citations on the face are the four below.

---

2. The four examiner-cited patent references (Google Patents "Citations (4)")

---

3. Reference-by-reference analysis

Threshold caveat on §102: To anticipate, a single reference must disclose every limitation of a given claim, arranged as in the claim (Net MoneyIN). None of the four references below contains a subject orientation-reference indicia read by a computer to re-image a computer-generated article and composite it with a live/stored subject image. That limitation is the core of independent claims 1, 13, and 20. Accordingly, on the face of the record none of these references anticipates an independent claim; the patent issued over all four. They are properly characterized as §103 obviousness references, and §102 anticipation exists only if one reads the claims to a broader scope than issued.

3.1 US3510210A — Von Haney / Xerox (filed 1967-12-15; published 1970-05-05)

Brief description: A live actor is televised (vidicon camera 22) and the signal is digitized and fed to a computer. The computer stores an animated character and "the desired dress patterns, hair styles and the like may be brought from computer storage 12 and superimposed on the figure in the computer output unit 26 in response to suitable external commands and/or programs." The disclosure expressly contemplates altering "clothes and hair" (e.g., reprogramming trousers to a skirt, short hair to long hair) and using "a small, finite number of luminous or reflective points on the actor's body" for tracking, with the body considered symmetric "with respect to a co-ordinate system with axes at the mid-lines of the body."

Potentially anticipatory subject matter (caveated):

• Claim 1 / Claim 20 (partially): teaches "computer means to generate a graphic image of an article" (dress patterns/hair) + imaging a subject + producing a combined (superimposed) output. But it lacks the "orientation reference means for each frame," the "re-imaging" of the article in accordance with that reference, and the "x‑y‑z axes + angular rotation" indicia. ⇒ No full anticipation of claim 1 or 20.
• Claim 10 ("plurality of articles for use by a human … fashion previewer"): reads loosely on clothes/hair selection, but claim 10 is dependent on claim 1 ⇒ not independently anticipated.
• Claim 23 ("alter the image of said article … keyboard/external commands"): the "external commands and/or programs" reprogramming is close; however the claim requires the base "computer means to generate said graphic image of the article," which claim 1 supplies ⇒ §102 fails, §103 support only.
• Claim 4 (record/transmit subject video to storage): only partially met.

3.2 US3621214A — Romney et al. (filed 1968-11-13; published 1971-11-16; 48 claims)

Brief description: Method/system for electronically generating perspective (shaded) images of three-dimensional objects on a two-space display. Data defining object surface portions is converted to a view plane, sorted to the display scan pattern, hidden surfaces eliminated, and the system "determin[es] which of the surface areas … are to be displayed according to the desired orientation of the object" (claim 3), modifying intensity accordingly. Repro duces the same 3-D object "resulting from different orientations relative to the viewpoint."

Potentially anticipatory subject matter (caveated):

• Claim 1 / Claim 20 (partially): supplies the "computer means to generate a graphic image of an article" and the re-imaging/orientation-dependent rendering concept that underpins the "x‑y‑z axes and angular rotation" element. But it is entirely silent on: a subject, subject orientation-reference indicia worn by the subject, subject video storage, frame selection, and compositing the re-imaged article onto the subject. ⇒ No anticipation of 1, 13, or 20.
• Claims 13 / 14 (re-imaging per orientation; continuous re-imaging): Romney teaches orientation-driven rendering, but not the garment-coded-indicia trigger of claim 13 or the subject-motion loop of claim 14.
• Net: strongest single-reference teaching of the "re-image the article to the orientation" element (relevant to an obviousness combination with Maloomian, below).

3.3 CA1020276A — Wilfred E. Ryan, "Identification of faces" (priority 1974-05-14; published 1977-11-01)

Brief description (LOW CONFIDENCE — content not retrieved): I could not retrieve the specification or claims of this Canadian patent from the sources available; I therefore cannot quote it. Based on the title and its placement as an examiner citation, it appears directed to identification/recognition of faces (facial-feature analysis), which would be topically relevant to the "subject orientation reference information" and "indicia" concepts (claims 1, 11, 12, 13, 15, 17, 20).

Potentially anticipatory subject matter: Cannot be responsibly mapped to specific claims without the text. Given the '585 claims require orientation indicia providing x‑y‑z axes and angular rotation used to re-image a computer-generated article (claims 1, 11, 12, 13, 15, 20), a mere face-identification disclosure would not reach those limitations. I flag this as an open item rather than assert a §102 mapping. Confidence: low.

3.4 EP0024862A2 — Harold Charles Taylor (priority 1979-09-04; published 1981-03-11)

Brief description: A video apparatus that synthesizes a stylised perspective image of a room whose regions A–G (walls, ceiling, floor, skirting, door, window) are each displayed in a user-selected colour (hue/saturation/black), so a user can judge proposed decorating schemes on a tri-colour CRT. Segment generation is by counter/slope-index logic and gating/switching (areas A–G). The stated purpose is to let "the effects of proposed decorating schemes … be appreciated and judged by the user."

Potentially anticipatory subject matter (caveated):

• Claim 1 / Claim 20 (partially): teaches "computer[synthetic]-means to generate a graphic image" and "means to display said … image," and user-selected change of appearance. But it lacks a subject, subject orientation-reference indicia, storage/selection of subject video frames, orientation-driven re-imaging, and compositing of article onto subject. ⇒ No anticipation of 1/20.
• Claim 21 ("alter the image of said article in a manner not theretofore defined"): the user colour selection is conceptually analogous, but claim 21 depends from claim 20's computer/video/orientation structure ⇒ not independently anticipated; §103 fodder only.
• Relevance to the specification's "inanimate setting / room / furniture" passage: EP0024862 is the closest art to the '585 specification's stated alternative use ("the subject … may be an inanimate setting, such as a room and the article be articles of furniture, so that a room may be redecorated in previewing"). It is a redecorating-preview system but uses a synthetic room, not a captured subject re-imaged with an article.

---

4. Prior art discussed in the '585 specification (not on the face of the patent, but relevant)

The specification's "Background and Discussion of the Prior Art" identifies these references; they are part of the intrinsic record:

Summary of the strongest combination (for a hypothetical §103 challenge): Maloomian (subject-image storage + frame selection + composite display) + Romney (orientation-driven computer re-imaging of a generated 3-D object) + Reiffel/La Russa/Matsuda (superimposition) would collectively cover most of claim 1 — but the orientation-reference indicia read to re-image the article (§ claims 1, 11, 12, 13, 15, 17, 18) is the point of novelty the applicant relied on, and it is not found in any of the above.

---

5. Bottom line

• US 4,539,585 was retrieved and confirmed; searched only under number 4539585.
• Four examiner citations: US3510210A, US3621214A, CA1020276A, EP0024862A2.
  • US3510210A (Haney/Xerox) — §102 candidate only against a very broadly read claim 1/20 (computer-generated "article" overlay on a televised subject); no full anticipation; §103 support for claims 1, 4, 10, 20, 21, 23.
  • US3621214A (Romney) — §102 candidate only for the orientation/re-imaging element standing alone; no full anticipation; key §103 reference for claims 1, 13, 14, 20.
  • CA1020276A (Ryan) — content not retrieved; low confidence; cannot responsibly map to specific claims; likely §103 relevance to the "indicia/orientation reference" concept at most.
  • EP0024862A2 (Taylor) — "generate a graphic image + display + user-alterable appearance"; no full anticipation; §103 support for claims 1, 20, 21 and the spec's room/furniture alternative.
• No single reference anticipates claims 1, 13, or 20; none of the four independently anticipates any dependent claim either, because each dependent claim inherits the un-met orientation-reference/re-imaging limitation of its base claim.
• Most material art overall is Maloomian US 4,261,012 (expressly discussed and distinguished in the '585 specification), especially in combination with Romney US 3,621,214.

Caveats / confidence: (1) I did not obtain the file wrapper/Office actions, so I cannot confirm the examiner's actual §102 vs. §103 rationale — the references are listed on the face without a rejection designation. (2) CA1020276A full text was not retrieved; treat §3.3 as unresolved. (3) U.S. patent numbering is reported here verbatim as cited (US3510210A, US3621214A; the specification's US 3,507,570 / 4,261,012 / 3,617,630 / 4,246,605 / 4,255,766), with no auto-correction.

I'll ground this obviousness analysis in the actual prior-art references of record, so let me pull authoritative detail on each cited reference and the background references named in the specification.

Now let me pull the background references the specification itself identifies, since those are also "prior art on this page."

Obviousness Analysis — U.S. Patent No. 4,539,585 ("Previewer")

35 U.S.C. § 103 (pre-AIA), assessed as of the 1981-07-10 effective filing date

---

0. Scope, premises, and a cross-reference note

This section builds on the two prior sections (bibliographic summary; litigation summary) and does not repeat them. Two cross-references:

• The prior sections established that no post-grant proceeding or infringement suit ever tested the '585 patent's validity. An IPR2013-00432 used Spackova as Ex. 1003 against another patent (Lennon's US 6,624,843) — that is a citation event, not a validity adjudication of '585. So the analysis below is a retrospective, hypothetical § 103 assessment, not a reconstruction of any actual Office action.
• The prior section flagged that claim 15 depends from claim 12 although it recites "orientation referring means" (garment subject matter). I treat that below as primarily a § 112 ¶ 4 problem, not a § 103 point — but it materially affects how much of claim 15 is even amenable to an obviousness analysis.

Date note (carried forward, not resolved): the prompt date is April 26, 2026; my earlier working date was September 27, 2026. This does not affect pre-AIA § 103 analysis of a patent that expired 2002-09-03.

---

1. Legal framework applied

Under Graham v. John Deere Co., 383 U.S. 1 (1966), and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007):

1. Determine the scope and content of the prior art;
2. Ascertain the differences between the prior art and the claims;
3. Resolve the level of ordinary skill in the art; and
4. Evaluate objective indicia (secondary considerations) if any are of record.

Because the effective filing date is July 10, 1981, pre-AIA § 103(a) governs and the "motivation to combine" must be supported by a rational underpinning (In re Kahn, 441 F.3d 977 (Fed. Cir. 2006)). KSR supplies the key additional tools here: (i) a known technique applied to a known device, ready for improvement, to yield predictable results; (ii) combination of prior art elements according to known methods; and (iii) where there is "a design need or market pressure" and a finite number of identified, predictable solutions, the combination is obvious to try.

§ 102 status of the four references of record (all appear on the face of the patent as "References Cited," so all were before the examiner):

The three pre-1978 references are unambiguously § 102(b) art. Taylor is the weakest of the four on prior-art status and should be relied on only for the peripheral (color/furnishing) claims.

---

2. Level of ordinary skill in the art (POSITA), ca. July 1981

A POSITA here would be a person with a bachelor's degree in electrical engineering or computer science (or equivalent) and roughly 2–4 years' experience in video systems and/or computer graphics/image processing, and would be familiar with:

• Video compositing and keying — chroma-key, video mixing, and half-silvered-mirror overlay were mature by the mid-1970s (all reflected in the '585's own Background discussion of Bourdier and Maloomian).
• Computer-graphics object transformation — rotation and translation of a stored 3-D model, projection to a 2-D view plane, and hidden-surface removal (the core of Romney).
• Marker/landmark tracking — digitizing a live video signal and recognizing a small array of high-contrast points (the core of Von Haney; also the core of Ryan's face identification).
• Conventional input devices — keyboards, joysticks, and switch banks for selecting and parameterizing images.

This is a system-integration art, not an art of unpredictable chemistry or biology. That matters: § 103 in a systems-art context is usually resolved on the strength of the articulated motivation, not on unpredictability of results.

---

3. Claim construction of the decisive limitations

Everything turns on four phrases.

(a) "computer means to generate a graphic image of an article to be previewed" (claim 1, 20). This is a means-plus-function element (§ 112 ¶ 6 equivalent under 1981 practice). The corresponding structure disclosed is "a computer graphic video image generator 20… designed to provide graphic images of an article, such as eyeglasses 21." So the function is: synthesize (not photograph) the product image. Von Haney discloses it; Romney discloses it.

(b) "subject orientation reference means for each frame" (claim 1) / "video image of a subject having subject orientation reference information" (claim 20). The disclosed structure is the equilateral triangle 15 worn by the subject. Von Haney discloses structural equivalents (a finite array of reflective/luminous body points) — discussed below.

(c) "means to re-image said article in accordance with said subject orientation reference means for the specific selected frame" (claim 1). This is the transform step: read pose → rotate/translate the synthesized article → re-render. Romney discloses the transform.

(d) THE PIVOT: "said orientation means comprising indicia providing x-y-z axes orientation and angular rotation orientation with respect to said axes" (claims 1, 13, 20; also claim 15). This "x-y-z + angular rotation" recitation appears in all three independent claims and is the most likely reason the application was allowed over the four references of record. Any serious § 103 case must therefore be built at this limitation — and it is where the case is both strongest (Von Haney's body-axis coordinate system) and most contestable (Von Haney used it for pattern-recognition simplification, not for article transformation).

---

4. What each reference of record actually teaches

US 3,510,210 — Von Haney (Xerox), "Computer Process Character Animation"

The most on-point reference, and remarkably so. From the specification and claims:

• A live actor 24 performs movements; a vidicon television camera 22 digitizes them and feeds a comparison unit 20 in computer 10.
• The actor's motion drives selection of stored figure positions and postures from computer storage 12, which are output at computer output 26.
• Express teaching of superimposing stored articles on the figure: "The desired dress patterns, hair styles and the like may be brought from computer storage 12 and superimposed on the figure in the computer output unit 26 in response to suitable external commands and/or programs."
• Express teaching of modifying the synthetic article: "the movements of a little boy may be changed to those of a little girl by reprogramming the trousers to a skirt and the short hair to long hair."
• Express teaching of a marker-based body coordinate system: "The otherwise difficult problem of pattern recognition in general is simplified in this case by the use of a small, finite number of luminous or reflective points on the actor's body. Further, the body of the actor and the light spot array may be considered symmetrical with respect to a co-ordinate system with axes at the mid-lines of the body."
• Express teaching of compositing and display: overlay of computer cells onto background art; output on CRT 28.

Source: US3510210A, Google Patents; patentimages PDF.

US 3,621,214 — Romney et al., "Electronically Generated Perspective Images"

• Generates perspective representations of three-space objects on a two-space display; "ordering the sets in relation to the two-space characteristics of the display."
• Claim 3: "determining which of the surface areas defined by the sets are to be displayed according to the desired orientation of the object."
• Claim 18: "changing said stored digital signal to correlate with a chosen orientation relative to the eyepoint from which the perspective view is generated; rearranging said changed digital signals to conform to the scanning pattern of the display."
• Claim 5: data representative of "locations, orientations and characteristics of surface portions."
• Quantizes objects into triangular surface primitives (see the related disclosure's "three overlapping planar triangles 26, 28 and 30").

Source: US3621214A, Google Patents; FPO.

This is the re-imaging engine. Romney teaches exactly the step Von Haney lacks: taking a stored model and re-rendering it correctly for an arbitrary orientation, with correct visibility. Romney is also the source for using a triangle as the geometric primitive — relevant to claims 11 and 15.

CA 1,020,276 — Ryan, "Identification of faces" (1977-11-01)

On the face of it, a face-identification reference. I was unable to retrieve its substantive text with the tools available (search step limit reached), and I will not fabricate its disclosure. What is verifiable and material:

• It is Canadian prior art published 1977-11-01 (over three years pre-filing).
• Google Patents lists US 4,539,585 among its "Cited By" — i.e., the examiner appears to have cited Ryan against Spackova, which is consistent with Ryan disclosing landmark/feature identification on a human face.
• Its title alone ("Identification of faces") supports an inference — which I state as an inference, not a verified disclosure — that it teaches identifying and locating features of a face for subsequent use, which is the antecedent of "identification means" (claims 17–18) and of placing reference indicia on a subject's head.

Source: CA1020276A, Google Patents.

EP 0 024 862 A2 — Taylor, "Video apparatus for visualizing effects of selected juxtaposed colours" (pub. 1981-03-11; US counterpart US 4,329,710)

• Generates a stylized perspective image of part of a room — areas A–E being first wall, ceiling, second wall, floor, skirting; optionally a door F or window with curtain G.
• The apparatus "enables each of the areas A–G to be displayed in any desired colour, whereby the effects of proposed decorating schemes may be appreciated and judged by the user."
• Uses colour composers 10 with manually operable hue/saturation/black controls; colour selectors 12 (switches) per area; a v.d.u.; and an alternative digital implementation storing "prescribed colours… as binary numbers in three memories 100, 102, 104."

Sources: EP0024862A2, Google Patents; US4329710, FPO.

Taylor is the art for the "furnishings"/room-redecorating embodiment and for the color-modification claims, and it independently shows a selectable, parameterized, computer-stored color scheme applied to a generated environment image.

---

5. References named in the '585 specification itself (supplementary art)

These are not in the "Citations (4)" list but the patent's own Background and Description expressly identifies them as prior art, which makes them fair game (an applicant's own characterization of the art is an admission):

• US 3,507,570 — Bourdier et al. (Audio Visuel France), "Composite optical display system." A semi-transparent input mirror 6 plus a projection screen and a slide projector 11, arranged so that projected slides of articles (spectacle frames, jewelry, make-up, clothes) appear on a plane coinciding with the virtual image plane of the viewer as formed by the mirror — i.e., the viewer sees himself wearing the projected article. That is precisely the half-silvered-mirror compositing (reference numeral 31) of the '585 and the "viewer" 32 of its FIG. 1. Source: US3507570A, Google Patents.
• US 4,261,012 — Maloomian. As characterized by the '585 itself: "a series of model posed images are video stored and a video composite image with the viewer's image is rendered on a television monitor." This is the video-mixer branch of claim 6 and the stored-frame branch of claims 1/4/5.
• The '585 also names Reiffel US 3,617,630; La Russa US 4,246,605; and Matsuda US 4,255,766 as teaching "a computer image generated display image and superimposing another image thereon." These are cumulative for the composite/display elements.

---

6. Combination A — Von Haney + Romney (+ Bourdier or Maloomian) → claims 1, 20 and their dependents

This is the central § 103 case.

Element-by-element mapping for claim 1

On the pivot limitation (x-y-z + angular rotation): Von Haney expressly locates the marker array within "a co-ordinate system with axes at the mid-lines of the body." A POSITA in 1981 knew that three non-collinear points on a rigid body uniquely determine that body's 3-D orientation — i.e., a translation and a rotation about the axes. Von Haney's "small, finite number of luminous or reflective points" is therefore a disclosure of pose measurement in six degrees of freedom, with the axes and rotations inherent in the disclosed coordinate frame. Romney supplies the mathematical apparatus for applying that orientation to a re-rendered object. This is a known technique (Romney) applied to a known device (Von Haney) to yield a predictable result — the KSR paradigm.

Claim 20 (transmission/real-time variant)

Claim 20 is broader than claim 1 in every respect except the same x-y-z/rotation limitation: "means to transmit a video image of a subject having subject orientation reference information," "means to process said subject video image," "computer means having means to input said subject orientation reference information and to re-image said article," "means to display said subject video image with said re-imaged article image as a composite." Every one of these is met by Von Haney's camera→digitizer→comparison-unit→output chain (Von Haney's "interface… involves the production of a video or television signal, converting that signal to a digital form"), and by Romney's projection/ordering/display chain. Claim 20 is, if anything, weaker than claim 1 because it drops the explicit stored-frame and frame-selection elements.

Dependent claims 2–12, 14, 16, 21–23

---

7. Combination B — Von Haney + Romney + Ryan → claims 13, 17, 18, 19 (the "virtual dressing room" claims)

Claim 13 requires: (i) a garment bearing a plurality of coded indicia; (ii) a video camera photographing the garment as worn; (iii) a digital image processor converting the codes into digital information instructions; (iv) a computer generating apparel images per those instructions; (v) display of the re-imaged apparel as if worn; and (vi) each coded indicia comprising orientation reference means with the x-y-z + angular rotation limitation; plus, in dependents, identification means (17), a binary reference number (18), and a form-fitting garment (19).

Motivation: the same one that drives the entire patent — eliminating the model/posing time and static-pose limitations the '585 itself concedes in its Background ("the need for pre-recording modelled images is minimized or eliminated"). A POSITA seeking to track a subject's whole body rather than just the head would naturally distribute the reference markers over the body — and putting them on a form-fitting garment is the ordinary way to do it.

---

8. Combination C — Taylor (EP 0 024 862) for the color and furnishing aspects

The '585's specification expressly extends the invention to "an inanimate setting, such as a room and the article be articles of furniture, so that a room may be redecorated in previewing," and claims 21/23 cover altering the article (including, per the description, its color). Taylor independently discloses previewing decorative color schemes on a generated room image, with digitally stored prescribed colors. Taylor thus meets the "furnishings/room" application and the color-modification concept, and reinforces the keyboard/manual-control element of claim 23. Its status as art (§ 102(a) rather than § 102(b)) makes it the least robust of the four, so it should be used only as corroboration, not as a primary reference.

---

9. Consolidated motivation to combine (the KSR "rational underpinning")

A prima facie case requires an articulated reason; the following are available on this record, in decreasing order of strength:

1. The primary reference itself suggests the combination. Von Haney expressly teaches bringing "dress patterns, hair styles and the like… from computer storage 12" and "superimpos[ing]" them on the figure, and reprogramming a garment (trousers→skirt) to change how the figure appears. That is the '585's core concept — the invention is suggested by the reference that the examiner cited against it. This is the single most damaging fact for the '585's claims 1, 20 and 21.
2. The reference supplies the solution to the exact deficiency the '585 admits in the prior art. The '585's Background states that Bourdier was "limited to a fixed physical position of the user" and that Maloomian was "limited to the particular number and particular poses that have been pre-recorded, and further requires substantial model or posing time." Romney teaches precisely the missing capability: generating a correctly oriented, visible-surface perspective view of a stored 3-D object for an arbitrary chosen orientation, thereby freeing the system from pre-recorded poses. The problem statement in the patent is the motivation.
3. Same field of endeavor / reasonably pertinent art. Von Haney (computer-generated figures driven by live video), Romney (computer-generated perspective objects), Bourdier and Maloomian (composite self-image-plus-article display) are all directed to combining synthetic imagery with live or recorded human imagery. All are analogous art (In re Bigio; In re Clay).
4. Known elements, known methods, predictable results. Camera → digitize → compare to stored model → transform model → composite → display. Every element and every interconnection was known; the combination yields no more than the expected sum of its parts.
5. Market/design pressure and a finite solution space. Retail eyewear and apparel selection was a recognized commercial need (the '585 admits the discomfort of trying on frames with non-prescription lenses). Given that need, a POSITA had a small, identified set of options: mirror-based optical compositing (Bourdier), video mixing (Maloomian), or digital overlay — and marker-based pose capture was the known technique for the tracking problem.
6. Explicit interchangeability of the compositing elements. The '585's own abstract recites that the images are "video mixed, optically mixed, or superimposed in digital image storage memory" — i.e., the applicant treated three known compositing techniques as substitutes for one another. Substitution of known equivalents is obvious.

---

10. Where the § 103 case is genuinely weak (candidate rebuttals)

An honest analysis must identify what a patent owner would argue, and where the record does not fully close the gap.

1. The "x-y-z axes + angular rotation" limitation is the real defense. Von Haney's coordinate system exists to simplify pattern recognition, and its axes are the body's mid-lines — not a frame devised to define the article's rotation. Romney re-orients an object relative to an eyepoint, not relative to markers worn by a live subject. Neither reference, taken alone, discloses deriving the article's 3-D orientation from reference indicia worn by the subject. The combination argument bridges the gap, but a patent owner would argue the bridge itself (measure subject pose → parametrically drive the object renderer) is the inventive contribution — and, importantly, none of the four references of record appears to disclose it expressly. This is the issue on which a § 103 challenge would likely stand or fall, and it is not free from doubt.
2. Ryan's disclosure is unverified. I could not retrieve CA 1,020,276's text with the available tools. Any argument keyed to Ryan — especially claims 17–18 (identification means / binary reference number) and the "identify the subject's features" element of claim 13 — is inference-based and must be corroborated before being relied on. I have flagged this rather than asserted it.
3. Real-time feasibility at 1981 processing rates. Claim 14 requires continuous re-imaging as the subject moves. Von Haney's pipeline was for frame-by-frame animation production, not interactive retail preview; Romney's shaded, hidden-surface-removed renders were not obviously real-time on 1981 hardware. A patent owner would argue this is a genuine, non-obvious engineering achievement. That said, claim 14 recites "means to process said variations… so that the re-imaged apparel is continuously displayed with subject movement" — a functional recitation not tied to any frame rate, latency, or hardware budget — which substantially undercuts the argument in the claim-as-written.
4. Claim 9 (head-shaped joystick) is an ornamental variant with little functional content — a weak claim for the patent owner, but also an unusual thing to make the centerpiece of an invalidity case; it is better handled as obviousness of a design choice.
5. Secondary considerations. I found no evidence of commercial success, industry praise, licensing, copying, or a nexus in the record available to me — and the patent expired for failure to pay maintenance fees in 2002 (per the prior section), with no litigation ever filed. A patent owner facing a § 103 challenge would have a difficult time assembling objective indicia, but the absence of such evidence is not itself evidence of obviousness; it simply leaves factors 1–4 to carry the case.
6. Claim 15's antecedent/independence defect is a § 112 issue, not a § 103 one. "The fashion previewer of claim 12" (which is an apparatus "previewer" claim, not a "fashion previewer" claim and which does not recite a garment) while reciting "each said orientation referring means" leaves claim 15 without a clear antecedent scope. As the prior section noted, I report the claim text as printed rather than correcting it. If claim 15 is indefinite, indefiniteness — not obviousness — is the cleaner ground.

---

11. Bottom line

A strong prima facie § 103 case exists against every independent claim and most dependent claims of US 4,539,585, built on:

• Primary: Von Haney US 3,510,210 — which alone discloses the purpose (superimposing stored dress patterns and hairstyles on a live-video figure and reprogramming garments), the tracking mechanism (a finite array of body markers in a mid-line-axis coordinate system), and the compositing output.
• Secondary: Romney US 3,621,214 — supplying the re-orientation/hidden-surface rendering of the stored article, and the triangular geometric primitive.
• Compositing alternatives: Maloomian US 4,261,012 (video mixer) and Bourdier US 3,507,570 (semi-transparent mirror) — the two alternatives the '585 itself treats as interchangeable.
• Ancillary: CA 1,020,276 (Ryan) for feature identification (unverified — flagged) and EP 0 024 862 (Taylor) for the color/furnishing preview and keyboard control.

The single limitation that resists clean anticipation-by-combination is the requirement, in claims 1, 13 and 20, that the orientation indicia provide x-y-z axes orientation and angular rotation orientation with respect to those axes — a limitation that appears to have been the point of novelty over the four references of record. The strongest invalidity position is therefore a KSR-style argument that (a) Von Haney's finite 3-D marker array in a body-axis frame inherently measures six-degree-of-freedom orientation, and (b) applying Romney's known object-reorientation technique to that measured pose was a predictable use of a known technique on a known device, undertaken in response to the very design need the patent itself identifies.

Confidence: high as to the disclosure content of US 3,510,210, US 3,621,214, EP 0 024 862 A2 and US 3,507,570 (verified from primary/patent-office sources cited above); lower as to CA 1,020,276 (text not retrieved) and as to US 4,261,012 (characterized only as the '585 itself describes it). I have also not reviewed the actual 1981–85 file history in the USPTO PatentCenter document viewer, so I cannot say whether the applicant overcame any specific rejection on the "x-y-z/rotation" ground — a step that would materially sharpen or weaken this analysis.