US 10009537

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

Let me dig into the litigation, IPR, and claim specifics.

Verification of the identifier

I searched for the literal number 10009537 / US 10,009,537 B2 and did not return or substitute any similar number. The record below is drawn from the authoritative Google Patents full text supplied for this patent, plus litigation/PTAB records I retrieved for that specific patent number.

---

1. Bibliographic record (per Google Patents, authoritative)

Priority-chain caveat (reported literally, not corrected): The specification's RELATED APPLICATIONS section states the chain as: continuation of Ser. No. 15/007,253 (filed Jan. 27, 2016) → continuation of Ser. No. 14/437,105 (filed Apr. 20, 2015; U.S. Pat. No. 9,338,348) → national phase of PCT/IL2013/050856 (filed Oct. 22, 2013) → U.S. Provisional 61/717,216 filed "Oct. 23, 2013" and 61/759,643 filed Feb. 1, 2013. The Google Patents front-page "priority date" is 2012-10-23, and the Samsung IPR expert declaration (Schonfeld, re U.S. 10,944,901) expressly assumes an earliest invention date of October 23, 2012. The specification's "2013" date for 61/717,216 therefore appears inconsistent with the front-page and PTAB records; I am flagging the discrepancy rather than resolving it.

2. Abstract (verbatim)

"A computerized method for computing the photo quality of a captured image in a device image acquisition system, comprising on-board combining of a plurality of quality indicators computed from said captured image and its previous image frames quality indicators and a confidence level for at least one of said quality indicators, and using a processor to determine, based on said combining, whether photo quality is acceptable and taking differential action depending on whether quality is or is not acceptable."

3. Family / prosecution chain

Per the published chain of the later continuation (US 2025/0159337 A1), the family runs:
14/437,105 (US 9,338,348) → 15/007,253 (US 9,661,226) → 15/582,722 (US 10,009,537) → 15/992,217 (US 10,659,682) → 16/867,919 (US 10,944,901) → 17/189,587 (US 11,252,325) → 17/666,585 (US 11,671,702) → 18/139,368 (US 12,250,452) → 19/021,306 (US 2025/0159337 A1). All share the same disclosure and the "Real time assessment of picture quality" title.

4. Plain-language claim overview — with an important limitation

Uncertainty flag (stated up front): The authoritative full text retrieved for US 10,009,537 contains the abstract, specification, drawings listing, and classifications but not the printed claims. I will therefore not reproduce verbatim claim language for this patent. The overview below is grounded in (a) the specification's sole summarized embodiment ("Embodiment 1" and its dependent Embodiments 2–13) and (b) claim sets appearing in the SnapAid-family PTAB/litigation record, which I could not confirm belong to the '537 patent specifically.

What the specification frames as the invention (Embodiment 1): a method for computing photo quality of a captured image in a device image acquisition system, comprising (i) computing at least one total quality indicator based on at least two quality indicators computed from the captured image and its previous image frames, to determine whether photo quality is acceptable, and (ii) providing corresponding detailed photo quality feedback.

Likely independent-claim structure, in plain language (from the family's asserted claim sets; confirm against the printed '537 claims before relying on it):

• A system claim — a device image-acquisition system (camera/lens module, image sensor, processor) that computes multiple quality indicators (QIs) characterizing quality from image data and/or hardware sensors (accelerometer, gyro, GPS), combines them (optionally with time-dependent confidence levels) into a total quality indicator, compares against a logical criterion / threshold, and either automatically actuates capture when the criterion is met and/or outputs a suggestion selected from a pre-stored table of suggestions while the criterion is not met.
• A method claim — combining a plurality of QIs computed from the captured image and its previous image frames, together with a confidence level for at least one QI (expressed in the family's papers as based on the probability factors P of the QI, calculated in the current and same-scene previous frames from "module 200" continuous video frames), and determining photo quality from that combination.
• Dependent subject matter in the family includes: 4D device coordinates; total-QI threshold vs. per-QI thresholds; OR-combinations of threshold conditions; automatic capture without a shutter press; changing lens/sensor module parameters based on the QIs; video quality quantification and video-improvement suggestions; blur-type detection alerts (object movement; "moving too fast in a vehicle"); and gesture-based shutter/delete (shake gesture). Whether each of these is an independent or dependent claim in the '537 patent is something I could not verify.

Note: The '537 claim set is not the same as its siblings — e.g., the PTAB record shows Samsung challenging claims 1–20 of US 10,944,901, claims 1–12 of US 12,250,452, and claim 1 of the '226, '682 and '537 patents being charted in invalidity contentions. So the '537 has at least a claim 1 that was specifically charted in Samsung's E.D. Tex. invalidity contentions.

5. Litigation and post-grant status (as of April 26, 2026)

• District court: SnapAid, Ltd. v. Samsung Electronics Co., Ltd. et al, No. 2:25-cv-00378 (E.D. Tex., Judge RWS/RSP), complaint filed April 10, 2025. Google Patents flags this patent as having "family litigation," and the Stanford NPE Litigation Database lists 2:25-cv-00378 as the case involving patent 10009537 (correcting a first-system, individual-inventor-started asserter). Eight SnapAid patents are asserted, with trial reportedly set for April 2027.
• PTAB: Samsung filed IPRs and at least one PGR against five of the eight asserted patents in late 2025 (e.g., IPR2025-01520, PGR2025-00083; petitions/exhibits at ptacts.uspto.gov petition IDs 1558466, 1558467, 1558471). In its patent-owner papers SnapAid argues Samsung "challenges only five of the eight asserted patents — those issued within the last six years," i.e., the older members of the family (including the '537) were not challenged. Samsung filed a Sotera+ stipulation regarding the five challenged patents.
• Reexamination: An ex parte reexamination, control no. 90/15,974, titled "REAL TIME ASSESSMENT OF PICTURE QUALITY," was filed Feb. 19, 2026, and the reexamination was ordered Mar. 10, 2026 (status: "Determination – Reexamination Ordered"). Uncertainty: the docket entry gives the title only, not the patent number; because the title is shared across the whole family, I cannot confirm which family member (e.g., the '537 vs. the '901 or '452) 90/15,974 covers. Do not treat this as confirmed for '537.

6. CAFC 2026 dockets — negative result

I searched 2026 Federal Circuit materials for this patent number and for "SnapAid." I found no Federal Circuit appeal, docket, or opinion involving US 10,009,537 (or any SnapAid patent) in 2026. The 2026 CAFC items returned (Centripetal v. Keysight, Ollnova v. ecobee, REGENXBIO v. Sarepta, Moskowitz v. Globus, NCS Multistage v. Nine Energy, Fortress Iron v. Digger Specialties, etc.) are unrelated. Note the timing: the E.D. Tex. trial is not until ~April 2027, and any PTAB Final Written Decisions from the 2025 filings would be later still, so an appeal of this patent to the Federal Circuit would not yet be ripe. Treat the absence of a CAFC record as a verified negative for this patent number as of the search date, not as proof no appellate activity exists anywhere in the family.

---

Items I did not verify (no fabrication)

1. The verbatim text of any claim of US 10,009,537 (claims section not present in the retrieved authoritative text).
2. Which specific patent is the subject of ex parte reexam 90/15,974.
3. Exact independent-claim count/boundaries for the '537 (e.g., whether claim 1 is a system claim and where the method claim begins).
4. The correct filing date of provisional 61/717,216 (specification says Oct. 23, 2013; front-page priority and PTAB record say Oct. 23, 2012).

Primary sources: https://patents.google.com/patent/US10009537/en · https://portal.unifiedpatents.com/litigation/Texas%20Eastern%20District%20Court/case/2%3A25-cv-00378 · https://npe.law.stanford.edu/patent/10009537 · https://ptacts.uspto.gov/ptacts/public-informations/petitions/1558467 · https://ptacts.uspto.gov/ptacts/public-informations/petitions/1558471 · https://sandbox.docketalarm.com/patentapps/US/90-15,974/ · https://www.patents-review.com/a/20250159337-real-time-assessment-picture-quality.html

I'll research litigation involving US Patent 10009537 using the sources you specified.

Let me verify additional details on the PTAB proceedings and check for any other cases.

Let me check for any appellate activity or additional defendants, and confirm the current district court status.

Litigation Involving U.S. Patent No. 10,009,537 ("Real Time Assessment of Picture Quality")

Bottom line: One (1) litigation is known — a single district court action in the Eastern District of Texas, filed April 10, 2025, by SnapAid, Ltd. against Samsung. The '537 patent is one of eight asserted patents in that case. There is no PTAB proceeding against the '537 patent specifically, and I found no other case (and no appeal) involving this patent.

---

Case 1 — SnapAid, Ltd. v. Samsung Electronics Co., Ltd. and [Samsung Electronics America, Inc.](/litigations/by-plaintiff/Samsung%20Electronics%20America%2C%20Inc.)

Role of the '537 patent in this case

• The complaint asserts eight related patents, all titled "Real Time Assessment of Picture Quality": the '348, '226, '537, '682, '325, '702, '901 and '452 patents.
• U.S. 10,009,537 is alleged to be infringed by Samsung's "Flaw Detection" feature (automatic post-capture analysis identifying motion blur, closed eyes, lens smudges, then prompting a retake) in Galaxy smartphones and tablets, e.g., the Samsung Galaxy S21 FE 5G.
• Accused claim elements mapped to the '537 patent in the complaint include: a system with an image capture component and processor; real-time computation of multiple quality indicators ('537 col. 2:19-25); automatic image capturing only after a first logical criterion predefined on said multiple quality indicators is satisfied (col. 4:10-13, cited at Compl. ¶127); a message provider selecting a suggestion from a pre-stored table (col. 4:15-20); and logical criteria predefined over a time-dependent confidence level (col. 3:12-14).
• Samsung's counterclaim: the Eleventh Counterclaim seeks a declaratory judgment that the '537 patent is invalid under 35 U.S.C. §§ 101, 102, 103 and/or 112.
• The complaint also includes patent-eligibility (§ 101) pre-emptive pleading for the '537 patent, asserting the claims are directed to a specific application (computing photo quality) and not an abstract idea.

Related PTAB proceedings (NOT directed to the '537 patent — included for context)

Samsung filed five petitions in September 2025 challenging only five of the eight asserted patents — the '537 patent was not among them:

• IPR2025-01519, IPR2025-01520, IPR2025-01521, IPR2025-01522 (all filed Sept. 5, 2025)
• PGR2025-00083 (filed Sept. 8, 2025)

Per the parties' filings, the three "unchallenged" patents are the '537, '348 and '226 patents. All five proceedings were resolved by discretionary denial (institution denied), with the Director's decisions issued on or about January 9, 2026. SnapAid had argued the petitions were a strategic, piecemeal attack on a single patent family filed after years of knowledge (2015 NDA disclosures under a Dual NDA dated Sept. 24, 2015), and Samsung countered with a "Sotera+" stipulation.

Party counsel of record (as reflected in the docket)

• SnapAid: James R. Nuttall (attorney in charge), Charles Everingham IV, Claire Abernathy Henry, James Winn, and Jacob Michalakes (pro hac vice).
• Samsung: Melissa Richards Smith.

---

Sources

• Google Patents patent page (flags the case as the first worldwide family litigation): https://patents.google.com/patent/[US10009537](/patent/US10009537)/en and https://portal.unifiedpatents.com/litigation/Texas%20Eastern%20District%20Court/case/2%3A25-cv-00378
• Stanford NPE Litigation Database, patent 10009537: https://npe.law.stanford.edu/patent/10009537
• Justia docket: https://dockets.justia.com/docket/texas/txedce/2:2025cv00378/[237215](/patent/237215)
• UniCourt docket summary: https://unicourt.com/case/pc-db5-casegu11f5f12148a8-[2076875](/patent/2076875)
• Complaint analysis (accused features, '537 mapping, counterclaims): https://ai-lab.exparte.com/case/dct/txed/2:25-cv-00378/doc/analysis/1
• Samsung counterclaims (incl. Eleventh Counterclaim, declaratory judgment of invalidity of the '537 patent): https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558467](/patent/1558467)/download-documents?artifactId=q7WPqI4ItbtkvGLDWSMIGyHAoTQaWwB7ALWo3TLnldJOLGzVoO1yB8Q
• SnapAid's discretionary-denial briefing identifying the '537, '348 and '226 patents as the three unchallenged patents and citing trial in April 2027: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1558469](/patent/1558469)/download-documents?artifactId=lfKpLi7tLnwnSitOrndDh3gGCbpEBiJzmRXCMKT9VmEu6aLkukFg1d0
• IP Verse PTAB list for Snapaid (all five Samsung proceedings = Discretionary Denial): https://ipverse.greyb.com/competitive-analysis/company/snapaid

---

Caveats and confidence notes

1. Literal identifiers preserved. The '537 patent = U.S. 10,009,537. I have kept the short-form "'537" citations exactly as they appear in the source documents and have not normalized or corrected any patent numbers, case numbers, or dates.
2. Outcome. There is no final outcome in 2:25-cv-00378; the case is mid-litigation (claim construction phase as of this date, with trial set for April 2027). Any statement of liability or validity would be speculation.
3. What I could not verify. My final search (for Federal Circuit / appellate activity arising from the January 2026 discretionary denials) returned no usable results before I exhausted my search budget. I therefore cannot affirmatively rule out a pending appeal of the PTAB discretionary-denial decisions. Based on the sources retrieved, no such appeal is documented.
4. On the PTAB proceedings. I want to be precise: the five Samsung IPR/PGR proceedings do not involve U.S. 10,009,537. They are listed only because they arise from the same eight-patent family asserted in the same district court case and are relevant to the litigation's procedural posture.
5. Status dates. Some docket aggregator pages (Justia, UniCourt) cache only through mid‑2025; the 2026 dates cited above (claim construction Oct. 22, 2026; trial April 2027; January 2026 institution denials) come from the parties' PTAB filings and the IP Verse PTAB table. Confirm against PACER for the most current docket.

I'll verify the structured block's default against live sources before concluding.

Samsung IPRs against the SnapAid family surfaced. I need to determine whether the '537 patent itself is among them.

Now let me pin down the full petition list and confirm whether '537 is among the challenged patents.

Proceedings overview

Zero AIA trial proceedings name U.S. Patent No. 10,009,537 as the challenged patent. The canonical structured block from the USPTO Open Data Portal reports no AIA trial proceedings, and my independent search of PTAB docket aggregators, the patent's Google Patents family/litigation record, and the litigation filings in SnapAid, Ltd. v. Samsung Electronics Co., Ltd., No. 2:25-cv-00378 (E.D. Tex.) confirms that — the only five PTAB cases in this patent family (four IPRs and one PGR, all filed by Samsung in September 2025) challenge five sibling patents, not the '537. The defensive posture this produces is therefore unusual and worth stating precisely: the '537 is neither invalidated nor hardened — it is PTAB-untested. No claim of the '537 has been canceled, no claim has been sustained in an AIA trial, and no petitioner has yet been estopped as to it. The patent owner is asserting claims 1, 2, 4–12, 14–20 of the '537 in E.D. Tex., and the only validity work product on it sits in Samsung's district-court invalidity contentions.

(Breakdown by status: active 0 / claims invalidated 0 / claims sustained 0 / settled 0 / institution denied 0 — no proceeding exists.)

---

US 10,009,537 — no proceedings to report

Applying the requested template literally: there is no {PROCEEDING_NUMBER} — {Petitioner} v. {Patent Owner} entry for this patent. I will not manufacture one. The default from the structured block holds after independent verification.

What I verified, and how: IP Verse's PTAB company record for SnapAid (last updated 2026-05-22) lists exactly five proceedings faced, all Samsung, all filed September 2025 — IPR2025-01519, IPR2025-01520, IPR2025-01521, IPR2025-01522 and PGR2025-00083. Similarly, SnapAid's own mandatory notices in PGR2025-00083 (filed 2025-10-21) enumerate the "open judicial or administrative matters" for the family and list only those five, each tied to a patent number that is not 10,009,537. The Stanford NPE Litigation Database entry for 10009537 shows a single case — 2:25-cv-00378 — with no PTAB overlay. (IP Verse SnapAid record; Stanford NPE DB – 10009537; PTAB E2E)

Why the '537 was left out — the "three unchallenged patents"

The '537 is one of eight patents SnapAid asserts against Samsung, and one of the three Samsung chose not to challenge. SnapAid's filings confirm the split: "[t]he district court action is moving forward without a stay... Samsung chose to challenge only five of the eight asserted patents in parallel IPR/PGR proceedings, leaving the three earliest-issued patents—with largely similar claims—uncontested at the PTAB" — and, separately, that Samsung "challenges only five of the eight asserted patents—those issued within the last six years—ignoring the others to avoid the settled-expectations problem."

The arithmetic lines up with the issue dates:

Adjacent family proceedings (flagged — they do not name the '537)

These matter to your defense strategy but are not proceedings on this patent. Reported here only so you are not blindsided by them; treat the status labels as second-hand pending verification.

• Claim-level detail where I could verify it (none of it is about the '537): IPR2025-01522 challenged claims 1–20 of the '682 under § 103 in three grounds — claims 1, 2, 5–8, 11, 14–16 over Anon (U.S. 8,508,622) + Takeuchi (US 2010/0149361) + Jasinski (US 2012/0201427); claims 17, 19, 20 adding Aisaka (US 2010/0246939); and claim 3 adding Alhadef. IPR2025-01521 challenged claims 1–20 of the '325 on the same Anon core with Kosaka, Aisaka, Garcia-Molina and Cheatle in the mix. IPR2025-01520 challenged the '702, whose claim 1 Samsung characterized as requiring weighting of QI1 (device motion), QI2 (under-/over-exposure), QI3 (calculated face properties) and QI4 (lens obstruction). PGR2025-00083 challenged the '452, with SnapAid arguing the '452 is not PGR-eligible because its effective filing date predates 2013-03-16. (IPR2025-01521 case record; IPR2025-01522 petition summary)
• Procedural posture: SnapAid filed requests for discretionary denial in each, arguing § 314(a)/Fintiv (no stay sought, trial set for April 2027 in E.D. Tex.; FWDs would land ~March 2027, one month before trial) plus § 325(d)/settled-expectations arguments built on a 2015 dual NDA and 2015–2017 portfolio disclosures to Samsung. Samsung countered with a Sotera+ stipulation agreeing not to pursue in E.D. Tex. any prior-art ground it raised or could have raised in the five proceedings — a stipulation that by its terms covers the five challenged patents only, not the '537.
• Estoppel consequence: because nothing was instituted, no § 315(e)(2) estoppel attaches to any of them — and Samsung's stipulation does not reach the '537.

Confidence caveats (stated rather than papered over): (i) I could not retrieve the underlying Board/Director decisions directly; the "discretionary denial" characterization comes from docket aggregators and is consistent with SnapAid's discretionary-denial campaign. (ii) The IPR2025-01521 docket shows some entries that are hard to reconcile with a denial — a "Request for Post-Institution Fees" (2026-03-04, petitioner) and "Notice: refund approved" (2026-03-11). Confirm the operative paper on PTAB E2E before relying on the status label. (iii) Whatever the outcome for the siblings, it is not a claim-level adjudication of the '537.

---

Strategic summary

Claim status on the '537: everything is UNTESTED. No claim — independent or dependent — has been canceled or confirmed by the PTAB. The asserted set per SnapAid's infringement position is claims 1, 2, 4–12, 14–20 (claims 3 and 13 are unasserted). Because I have no FWD for this patent, I cannot tell you which independent claim carries the case; the '537 issued 2018-06-26 from Application No. 15/582,722 (continuation of 15/007,253 → 14/437,105, now U.S. 9,338,348 → PCT/IL2013/050856), with Google Patents listing a 2012-10-23 priority date and a 2033-10-22 anticipated expiration — a long runway, which is precisely why this patent is worth defending carefully rather than dismissing.

Estoppel landscape. This is the rare case where estoppel helps you less than usual and hurts you less than usual at the same time. Samsung's IPR/PGR filings produced no institution, so no § 315(e)(2) estoppel bars Samsung from re-running prior art; and the Sotera+ stipulation Samsung gave the court is expressly limited to the five challenged patents. For the '537, that means the full universe of prior art remains live in district court — including the references Samsung charted as Exhibits 537-Alhadef through 537-Wu in its invalidity contentions, and including system/printed-publication art that would be awkward in an IPR. If you are a different defendant, you are in the same position: nothing constrains you, and nothing has been decided. You will also have to reckon with SnapAid's priority argument — Samsung contends no asserted '537 claim is entitled to a date earlier than 2013-10-22 (while reserving the position that the '537 may not even be entitled to that), which is a live § 102/§ 103 battleground rather than a settled one.

Pattern signals. One petitioner, one filing date cluster (2025-09-05/08), five parallel challenges, all denied. Samsung is a serial PTAB filer (~1,120 PTAB cases) but selected this family surgically, avoiding pre-2019 issuances. SnapAid is a one-case, individual-inventor-started NPE-style plaintiff (Dr. Ishay Sivan) with a single district-court assertion and no PTAB appeals in the family — there is no defensive aggregator in this chain; the Unified Patents entry on the Google Patents page is merely their litigation dataset recording 2:25-cv-00378, not a Unified-filed petition. The pattern that should worry you most: SnapAid is now 2-for-2 (plus 3) on discretionary-denial advocacy in this very family. If you file an IPR on the '537 after being sued in E.D. Tex. with trial set for April 2027, expect a § 314(a) Fintiv/settled-expectations fight that you may well lose.

---

Recommended next steps

1. Do not describe this patent as "instituted" or "invalidated" anywhere in a brief or a client memo. There is no FWD, no institution decision, and no appeal to link to for the '537. Cite the ODP result and the PTAB E2E docket showing the five Samsung proceedings are directed to '901, '702, '325, '682 and '452 respectively.
2. Pull Samsung's '537 invalidity contentions as your free prior-art search. The charted references (Exhibits 537-Alhadef through 537-Wu) and Samsung's priority challenge (no asserted claim earlier than 2013-10-22) are the best public validity workup on this patent. The documents surfaced via the PTAB petition-file mirrors at PTAB E2E and are quoted in SnapAid's own papers.
3. Bank on the district court, not the Board. With no estoppel and a hostile discretionary-denial climate in this family, an E.D. Tex.–anchored invalidity case (including § 112 and § 101 theories, which are outside IPR scope entirely) is your realistic path. If you do file an IPR, file it before being served in a parallel suit, and be prepared to distinguish SnapAid's settled-expectations record — note that the 2015 NDA and portfolio disclosures were to Samsung, so a non-Samsung petitioner has a materially cleaner § 314(a) story.
4. Track the April 2027 E.D. Tex. trial date and the sibling proceedings' disposition. If any Samsung petition was in fact instituted rather than denied (see the fee-entry anomaly in IPR2025-01521), the resulting claim-construction and validity findings on the shared specification are worth monitoring, even though they cannot collaterally estop you on the '537.
5. Watch for a new petition window. Today is 2026-09-25; the '537 is asserted and untested, so a fresh filing remains possible from Samsung or a co-defendant. There is no statutory bar to a new petition on the '537, but any petitioner sitting on notice of the family since 2015–2025 should expect SnapAid to re-run the same settled-expectations playbook that worked on the siblings.

Bottom line for a defendant: the absence of PTAB activity here is not a sign the patent is weak or unasserted — it is a deliberate targeting choice by the only defendant in the case. You are facing a live, unreviewed, asserted patent with a 2033 expiry, no adverse PTAB record to exploit, and no petitioner-friendly estoppel or FWD to borrow. Build the invalidity case in the district court, and treat the Samsung contentions as your starting point rather than your ceiling.

I'll research the assignment record for US 10,009,537, starting with the USPTO assignment data and any successor/transfer records.

Let me dig deeper into the recorded assignment specifics and the patent's prosecution counsel.

Let me check for a possible separate IP-holding entity and try to pin down the recorded reel/frame.

I hit the tool-call limit before I could pull the raw Assignment Center record itself, so the reel/frame numbers below are not retrieved. Everything else is grounded in the sources actually returned.

---

Ownership reconstruction — US 10,009,537 B2 ("Real time assessment of picture quality")

Source note / limitation up front: I could not open the USPTO Patent Assignment Center record (assignmentcenter.uspto.gov / assignment.uspto.gov) or the Google Patents "Legal Events" table directly within this session. The only assignment event I can substantiate for the '537 is the one mirrored in the authoritative Google Patents text supplied for this patent, plus corroborating file-history artifacts produced as PTAB exhibits. No reel/frame number is stated below because I did not retrieve one. I will not guess one.

---

Inventors

Filing-date discrepancy — now resolved. My earlier section flagged a contradiction between the specification's statement that provisional 61/717,216 was filed "Oct. 23, 2013" and the front-page/PTAB record of 2012-10-23. The PTAB Patent Owner Exhibit List returned in this search resolves it: Exhibit 2005 = "U.S. Provisional Application No. 61/717,216, filed Oct. 23, 2012" and Exhibit 2006 = "61/759,643, filed Feb. 1, 2013." The 2012 date is correct; the specification's "2013" is a typographical error in the patent. Also note the Union Patents portal lists the '226 priority as 2012-10-22 (PCT filing date) rather than 10-23 — a one-day discrepancy I am reporting, not correcting.

Unusual-pattern assessment: the inverse of the classic warning sign. The standard red flag is "all inventors depart the original assignee within 12 months of filing." Here the sole inventor never left — he is the assignor, the original assignee's founder/CEO, and the principal behind the asserting plaintiff. That is the signature of an individual-inventor-started portfolio, not a corporate fire-sale. There is no second inventor to depart, and no evidence of any inventor-side transfer other than the single company assignment below.

---

Original assignee

SNAPAID LTD. (Israel) — named as assignee on the issued '537. Address as recorded varies across the family: Tel Aviv on the '537 and '901; Ness Ziona on the '452 (issued 2025). The 2015 fee transmittal for the parent 14/437,105 gives "SNAPAID LTD., TEL AVIV, ISRAEL 6935635."

• Did it ship a product? Yes, in a limited sense — a real, released SnapAid Android application. PTAB Exhibit 2012 is the "SnapAid Android App User Manual," attached to an October 8, 2015 email from Sivan to Samsung; Exhibit 2014 is a "SnapAid Manufacture Presentation." The company's own portfolio summary (Ex. 2019, 2017) markets five technology lines including "catching the best non-blurred picture with confidence estimation (patented)" and head-pose estimation. So the operating story is a small Israeli software startup seeking to license ODM/OEM camera modules, not a mass-market hardware vendor.
• Primary line of business: mobile computational photography / image-quality software and licensing.
• Current status: No bankruptcy, dissolution, or acquisition found. SnapAid Ltd. still exists and is the named plaintiff in SnapAid, Ltd. v. Samsung Electronics Co., Ltd., No. 2:25-cv-00378 (E.D. Tex., filed 2025-04-10). However, the third-party startup profile returned describes SnapAid as "Inactive" with "approximately 1" person on record, and its app is long off the market. Treat "dormant-but-not-dissolved" as the working characterization, with the caveat that a startup-aggregator profile is a weak source.

---

Assignment timeline

One recorded assignment surfaced for this patent. It is the same-day inventor-to-company assignment reflected in the Google Patents legal-events feed for the '537:

• 2017-04-30 (executed) / recorded 2017-04-30 — Reel not retrieved/Frame not retrieved
  • Conveyance: Assignment of assignors' interest (the legal-events entry reads: "SNAPAID LTD. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SIVAN, Ishay")
  • Assignor: SIVAN, Ishay
  • Assignee: SNAPAID LTD.
  • Correspondent: not retrieved. Flag: the § 74 attorney/agent of record printed on the sibling family members is May Patents Ltd. (see US 10,944,901 and US 12,250,452). I could not confirm that May Patents Ltd. was also the correspondent on the assignment recording, so I am not treating it as a finding — only as the name to check first.
  • Context: inventor-to-company title perfection, executed and recorded on the same day the '537 continuation (Ser. No. 15/582,722) was filed. Routine startup housekeeping, not a transfer to a third party.

One earlier recordation in the priority chain (parent application — not a '537 event):

• The filing-fee transmittal for the parent 14/437,105 (signed January 20, 2016, attorney docket SIVAN-001-US, signature /yb/) requests that assignee data "SNAPAID LTD., TEL AVIV" be printed on the patent and states that "the document has been filed for recordation as set forth in 37 CFR 3.11." This evidences an earlier Sivan → SnapAid Ltd. assignment recorded against the parent. Reel/frame not retrieved.

No post-issuance assignments. No later assignee change appears for the '537 or for any of the eight asserted family members — every one still lists SnapAid Ltd. as assignee, and SnapAid sued in its own name rather than through a holding subsidiary. No security agreement, lien, merger, change-of-name, or release recording was surfaced. I flag these as absences in the sources retrieved, not as proof that no such record exists — the reel/frame lookup is exactly what would settle it.

Isolated data point worth diligence (not a '537 transfer): an IEEE/1790 Analytics citation listing for the related head-pose patent (filed 2020-02-20, granted 2021-05-25) names the assignee as "SNAP-AID PATENTS LTD" rather than SnapAid Ltd. If accurate, that is a distinct IP-holding affiliate within the same founder's orbit — and precisely the structure one would want to check for the camera-quality family. I found it in a single source and could not corroborate it; do not treat it as established.

---

Timeline diagram

timeline
    title Ownership of US 10009537
    2012 : Sivan files first provisional
    2013 : PCT application filed
         : Second provisional filed
    2015 : National phase enters US
         : Samsung dual NDA executed
    2016 : Parent fee transmittal names SnapAid
    2017 : US 15582722 filed
         : Sivan assigns rights to SnapAid Ltd
         : Assignment recorded same day
    2018 : US 10009537 granted
         : Further continuations filed
    2025 : SnapAid sues Samsung in Texas

---

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only recorded transfer is inventor → operating company (SNAPAID LTD., recorded 2017-04-30). Title has never left SnapAid Ltd.: the assignee of record on the '537 and on all eight asserted continuations is SnapAid Ltd., and the plaintiff sued in that name. No "IP / Holdings / Ventures / Licensing" shell appears anywhere in the chain.

2. Known asserter in the chain — PRESENT (weak-to-moderate). SnapAid does not appear on the enumerated high-frequency lists (Acacia, Marathon, IV, IPNav, Wi-LAN/Mosaid/Conversant, Vringo, Pendrell, Round Rock, Spangenberg et al.). But it is in the Stanford NPE Litigation Database for patent 10009537, categorized as "Individual-inventor-started" (5), which is an NPE taxonomy entry, and RPX/Unified surface the case as the "first worldwide family litigation." So: on a public NPE list, yes; a repeat-player asserters, no — 2:25-cv-00378 is its first and (so far) only suit.

3. Repeat correspondent across the chain — UNCLEAR. There is only one link in the chain, so recurrence cannot be tested within this patent. Worth noting for the family: May Patents Ltd. is the attorney of record on the issued siblings ('901, '452), and the prosecution docket is a self-referential "SIVAN-001-US" with a bare /yb/ signature — consistent with a very small, founder-controlled filing shop. I have not verified May Patents Ltd. as the assignment correspondent, so this is a lead, not a finding.

4. Cascading transfers — NOT PRESENT. The opposite: zero post-issuance transfers across nine family members filed between 2017-04-30 and 2025-01-15. Nine continuations, one owner.

5. Pre-litigation transfer — NOT PRESENT. No assignment at all in the ~8 years preceding the 2025-04-10 complaint, let alone within 6 months. The plaintiff is the original assignee, so there was no standing/venue clean-up transfer to find.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11, receivership, or asset-sale record surfaced for SnapAid Ltd. The company is described as inactive/dormant, not insolvent.

7. Privateering — NOT PRESENT (inverse configuration). Privateering is an operating company parking patents with an NPE to assert against competitors. Here the same entity that developed and demoed the technology (SnapAid) is the asserter, against the large company (Samsung) it once tried to partner with — after a 2015 NDA exchange, a 2017 purchase/license inquiry, and Samsung's refusal, per SnapAid's own PTAB briefing and Exs. 2007–2019.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. SnapAid remains the owner and is affirmatively asserting.

---

Verdict

NPE — moderate confidence.

The assignment record itself shows none of the classic troll-chain indicia: a single same-day inventor→company assignment (2017-04-30, SNAPAID LTD.), no shell transfer, no cascading LLCs, no pre-suit assignment, no bankruptcy sale, and no security interest — and the asserted plaintiff is the original assignee rather than a downstream acquirer. What tips the call to NPE is behavioral and taxonomic rather than structural: Stanford's NPE Litigation Database classifies SnapAid under "Individual-inventor-started" for this exact patent, and the owner is a dormant, ~1-person entity that no longer ships product yet is asserting eight patents against Samsung in E.D. Tex. — not an operating company suing a competitor over products it currently sells. Source for verification: USPTO Assignment Center (search patent 10009537; the reel/frame I could not retrieve is the item that would finalize this) and Google Patents legal events for US 10,009,537.

I'll search for the authoritative record of this patent's cited references and verify each.

Let me pull the front-page citations and the family's PTAB prior-art record directly.

Prior-Art Analysis — U.S. Patent No. 10,009,537 B2

Analysis date: April 26, 2026 · Patent: US 10,009,537 B2, "Real time assessment of picture quality," App. No. 15/582,722, granted 2018-06-26, sole inventor Ishay Sivan, assignee SnapAid Ltd.

---

0. Scope, sources, and what I could and could not retrieve

Identifier screening (literal-compliance note). Searching the number 10009537 returns several unrelated documents, and I have not auto-corrected or conflated them: DE10009537A1 (German — authenticating a mobile radio terminal via a protocol conversion unit) and EP10009537 (the application number of EP2256143A1, "Single solvent polymer extraction methods," Metabollix/METABOLIX). Neither has anything to do with US 10,009,537. Everything below concerns US 10,009,537 B2 only.

Source limitation — state this up front. The authoritative full text supplied for US 10,009,537 contains the abstract, specification, drawing listing, and classifications, but not the printed front page's "(56) References Cited" list. I attempted to retrieve that list from Google Patents, FreePatentsOnline and Justia and did not succeed within this session, and I could not open USPTO Patent Center / Public PAIR directly. Accordingly, the references analyzed below are the citations that appear on the face of the specification itself — i.e., the "Conventional technology constituting background…" list in the Background section and the in-text citations in the Detailed Description. This is a real gap: the Examiner-cited art in box (56) of the printed patent may add references I have not seen. Pull the printed '537 (or its PDF) before filing anything.

What is solid. The claim set is established (recovered in the prior section from two corroborating sources — Justia's reproduction and SnapAid's PTAB papers quoting "'537 Patent at 22:8–33"). Claim 1 is a method claim; claim 10 is a single-enclosure device claim; claims 2–9 depend from 1 and 11–20 depend from 10. The asserted set is claims 1, 2, 4–12, 14–20.

---

1. The § 102 framework I am applying (and why it changes the answer)

Anticipation under § 102 requires a single reference disclosing every limitation, arranged as in the claim. Because every dependent claim of the '537 incorporates claim 1 or claim 10, a reference that fails to disclose the independent claim cannot anticipate any dependent claim either.

That produces a blunt but correct first-order result:

On the record I could verify, no cited reference discloses all limitations of claim 1 or claim 10, and therefore no cited reference anticipates any claim of the '537. The case is a § 103 combination case (developed in the earlier Obviousness section), not a § 102 case.

The single limitation that defeats anticipation across the board is claim 1's final element — "where second weight (c2) is partially based respective values of previous images in said video" — together with the express weighting elements (c1, c2). Below I give, for each cited reference, the closest claim(s) it bears on and exactly what is missing, so you can see where a § 102 theory might be salvaged if a reference's full text discloses more than its abstract suggests.

Prior-art status is date-sensitive. If the '537 is entitled to the 2012-10-23 provisional (61/717,216), pre-AIA § 102 governs and § 102(b) art is anything published before 2011-10-23. If only the 2013-10-22 PCT date (PCT/IL2013/050856) is available, AIA § 102 governs. Samsung contends no asserted '537 claim is entitled earlier than 2013-10-22. This is live and unresolved — and it is outcome-determinative for US 2013/0155474, the most on-point reference (published 2013-06-20, i.e., after 2012-10-23 and before 2013-10-22).

---

2. The citation universe on the face of the '537

2.1 Patent documents expressly cited in the Background as "conventional technology"

2.2 Patent documents cited in the body of the Detailed Description

2.3 Non-patent literature cited

---

3. Reference-by-reference analysis

For each reference: full citation · dates · brief description · § 102 assessment (which claim(s), if any, it could anticipate). Dates not verified in this session are flagged; treat the printed patent's (56) box as the authoritative citation list.

---

R1 — US 2013/0155474 A1 — "Systems and methods for automatic image capture on a mobile device"

Dates. Published 2013-06-20. Filing/priority date not verified in this session (this is the critical unknown).
Description (verified from Google Patents). A mobile device evaluates image quality in real time, before capture, by measuring parameters including angle, motion and distance of the device; "[p]arameter values may be combined into a group threshold or total overall quality score" which must be "exceed[ed]… before the image can be captured"; the system gives feedback instructions to the user (e.g., "hold the camera steadier") and expressly addresses "blurriness… motion blur caused by the user moving the camera."
§ 102 assessment.

• Closest claims: claim 1 (and its dependents 2, 5). R1 discloses a pre-capture quality evaluation on a mobile camera, a motion parameter value, combination into a total overall quality score, a threshold comparison, and responsive action (feedback / capture gating). That maps much of claim 1 elements (a)–(j) and claim 2 (affecting the camera, via capture gating) and claim 5 (alerting the user).
• Fatal gap for anticipation: R1 (i) frames the combination disjunctively ("a group threshold or total overall quality score") and does not expressly disclose the weighted total (c1·QI1 + c2·QI2) required by claim 1(g); and (ii) discloses nothing about the second weight being based on previous images (claim 1(k)). Both are required. Therefore R1 does not anticipate claim 1 or any claim depending from it.
• Priority caveat (decisive). R1 published 2013-06-20 — after the 2012-10-23 provisional date and before the 2013-10-22 PCT date. If the 2012-10-23 date holds, R1 is not § 102 prior art at all (it is not even § 103 art, absent pre-AIA § 102(e) status from an earlier effective filing date, which I could not confirm). If only 2013-10-22 holds, R1 becomes available. This one reference's date controls whether Ground 1 in the Obviousness section survives.
• Verdict: § 103 material, not § 102. Anticipates no claim.

---

R2 — WO 2006/040761 A1 / US 2007/0195174 A1

Dates. Reported WO 2006-04-20; US 2007-08-23. Not re-verified.
Description. Known to me only through the '537's own characterization, which is itself an applicant admission: the interface module enables defining "scene dynamics of the captured image," where scene dynamics include "image motion speed and motion speed of the subjects," and enables "setting the captured image attributes relative weight for the compution of the total image grade."
§ 102 assessment.

• Closest claims: claim 1. As characterized, R2 supplies the express weighting element (c1, c2) toward a total image grade, and a subject-motion characteristic — i.e., claim 1(c), (f), (g) and part of (d).
• Gap: no disclosure of object recognition per se as claimed (claim 1(d) is disjunctive — "objects in, or one or more characteristics of, the image" — so "motion speed of the subjects" may satisfy it), and critically no disclosure of c2 based on previous frames (claim 1(k)).
• Verdict: § 103 material (and admissible as an applicant admission). Anticipates no claim, because of claim 1(k).

---

R3 — US 2009/0278958 A1

Dates. Published 2009-11-12 (reported). Not re-verified.
Description. Again known to me only through the '537's own quotation — an applicant admission: "The scoring of a current base image may be based on scores which have been given to previously captured base images. In such a manner, redundant calculation may be avoided."
§ 102 assessment.

• This is the single most dangerous reference on the face of the '537, because it is the only cited item that speaks directly to claim 1's final element (k) — the one limitation that defeats every other reference.
• Closest claims: claim 1 (and, if its full text goes further, claim 7 / claim 17's adaptive threshold). If R3 discloses not merely scoring a current image from prior images' scores but also a weighted, total quality value compared to a threshold with responsive action, it could present a serious § 102 question against claim 1.
• Gap as I can currently verify: R3 is quoted for scoring (a QI), not for weighting (c2), not for a total value, and not for the threshold/action steps. On the face of the '537's own quotation, R3 does not disclose every limitation, so it does not anticipate.
• Action item: this is the one reference whose full text I would pull first. I could not verify it, and I flag it as a potential § 102 reference pending review — I will not assert either way.
• Verdict: § 103 material (near-certain); potential § 102 for claim 1 pending full-text verification.

---

R4 — U.S. Pat. No. 7,362,354 B2 (Lin) and family WO 2003/069559 / US 2003/0151674 A1

Dates. US 2008-04-22 (reported, partly verified); WO family 2003.
Description (verified in substance). In-camera assessment of scene quality: computes a figure of merit for each of three photo-quality attributes — (a) photo sharpness, (b) face quality, (c) presence/absence of flare ("flare" being regional over/under-exposure); "compares each figure of merit to a respective pre-set threshold"; and "provides photo quality feedback." Uses face detection (RGB → L*a*b*, red-eye/binary mask) and regional pixel-percentage flare thresholds (≥50%).
§ 102 assessment.

• Closest claims: claim 1 (elements (d), (e)); claim 5 (alerting a user); and, on the over/under-exposure subject matter, nothing in the granted claim set as I recovered it.
• Gap: R4's scheme is per-attribute thresholding, not a weighted total (claim 1(g)); no c2 from prior frames (claim 1(k)); no automatic capture gating. It therefore cannot anticipate claim 1, and by dependency cannot anticipate claims 2–9 or 11–20.
• Verdict: § 103 material; supplies the "detecting or recognizing objects" (face) element if a broader reading is needed. Anticipates no claim.

---

R5–R11, R14–R15 — the references I could not verify

R5 US 7,688,379 · R6 US 2009/0263028 · R7 US 2006/0017820 · R8 US 2006/0056835 · R9 US 8,125,557 · R10 "WO201114812" · R11 US 7,920,727 · R14 US 2011/0096180 · R15 US 2013/0188866

Dates. Listed above; not verified in this session.
Description. These appear in the '537's Background/Description as undifferentiated "other systems," with no one-line characterization for R5–R11 (unlike R1, R2, R3, which the applicant characterized). R14 and R15 are cited by the patent for DCT/Fourier blur testing and chroma/exposure aesthetic modelling respectively.
§ 102 assessment.

• I did not retrieve their contents and do not assert what they teach. To do otherwise would be fabrication.
• What I can say is structural: they are cited only as background/technique references, and R14/R15 are cited for single QI-computation techniques (blur metric; aesthetic model). A reference relied on for one QI computation technique cannot, without more, disclose the combination of claim 1 (weighted total + threshold + action + previous-frame c2). None of these can be asserted as an anticipation reference on the present record.
• R10 deserves a specific flag: the printed identifier is "WO201114812", and I have reproduced it literally. It is ambiguous on its face (WO 2011/14812 vs. WO 2011/148121 are different documents). Do not resolve it by assumption; retrieve the actual number from the printed patent.
• Verdict: unknown; not citable as anticipation.

---

R12 — U.S. Pat. No. 8,040,382 B2 (Kahn et al.) — "Method and apparatus for improving photo image quality"

Dates. Issued 2011-10-18 (reported, verified in substance).
Description (verified). A camera with a motion sensor — "in one embodiment… an accelerometer" — predicts local minima of hand tremor; "overlay logic to overlay the points of minimum motion with a frame sync signal" and "shutter logic to activate an image capture mechanism at a frame sync point that is close to the point of minimum motion"; a global motion vector calculator with logic to "identify a frame having minimal global motion," to "identify a blurry frame, remove the blurry frame, and replace the blurry frame with a frame interpolated from adjacent frames"; preview mode; motion data stored as metadata associated with the image; shutter triggered by a proximity sensor.
§ 102 assessment.

• Closest claims: claim 10 and claim 20. This is the best single reference in the citation universe on the claim 10/20 subject matter: accelerometer-based motion sensing, capture triggered on a motion criterion, image storing logic. Claim 20's "accelerometer, a gyroscope, a GPS, or a step counter" recitation is squarely met as to the accelerometer alternative.
• Gap: claim 10 additionally requires QI2 (an image-analysis value) with its weight c2, a weighted total, a threshold comparison, and — via claim 1's incorporation of (k) — c2 based on previous images. R12 supplies frame-adjacency (aggregating/replacing adjacent frames' motion data) but not a weighted total of QI1·c1 + QI2·c2 compared to a threshold. It does not anticipate claim 10, and therefore cannot anticipate claims 11–20.
• Verdict: § 103 material (strong); anticipates no claim.

---

N1–N3 — non-patent literature

Dates. Ke/Tang/Jing 2006; Liu/Chen/Wolf/Cohen-Or 2010; Hartley/Zisserman 2004. All comfortably pre-date any priority theory.
Description. N1 = high-level-feature photo-quality assessment (no-reference and full-reference); N2 = computational evaluation of composition rules (rule of thirds); N3 = multi-view geometry for 3D reconstruction.
§ 102 assessment. Each is a printed publication that qualifies under § 102(a)(1)/pre-AIA § 102(b). But each addresses one aspect of the '537 (quality-feature scoring; composition; 3D scene reconstruction) and none discloses a weighted total quality indicator with previous-frame-dependent weights compared to a threshold with responsive action. They anticipate no claim; N1 is nonetheless the anchor for the "full reference / no reference" quality-scoring element and is useful § 103 corroboration of POSITA knowledge.

---

4. Summary matrix

---

5. Conclusions

1. No cited reference anticipates any claim of US 10,009,537 on the record I could verify. Every claim at issue depends from claim 1 or claim 10, and both independent claims require a weighted total value (c1·QI1 + c2·QI2) compared against a threshold, with claim 1's final element requiring c2 to be based on values of previous images. No single reference in the citation universe discloses that combination.

2. The two references worth pulling in full before finalizing any validity position are:

   • US 2009/0278958 A1 (R3) — the only cited reference that speaks to the previous-frames element (claim 1(k)). It is quoted verbatim in the '537 itself, so it is an applicant admission and prior art regardless of whether the Examiner listed it. If its full text goes beyond mere "scoring," it is the only realistic § 102 candidate against claim 1.
   • US 2013/0155474 A1 (R1) — the closest reference overall on claim 1 elements (a)–(j), but its 2013-06-20 publication date straddles the priority dispute: valid art only if the '537 is limited to the 2013-10-22 PCT date. Its filing/priority date — which I could not verify — decides whether it is § 102/§ 103 art at all.

3. R12 (US 8,040,382 B2) is the strongest single reference against the claim 10 / claim 20 subject matter (accelerometer/IMU motion sensing at capture, capture gating, image storage), but it supplies no weighted total and is therefore § 103, not § 102.

4. Applicant-admission references carry unusual weight here. R2 (WO 2006/040761) and R3 (US 2009/0278958) are characterized in the '537's own Background, and the specification also admits that weighting was known ("the weight is either constant or can be change by the user manually") and that error/probability estimation is known (Kalman filters, pattern-recognition probability). These admissions are usable against the claims irrespective of the printed (56) list.

5. Gaps I will not paper over:

   • The printed front page "(56) References Cited" list was not retrieved in this session; Examiner-cited art may exist that is not analyzed here.
   • R5–R11, R14–R15 contents were not verified; I make no assertion about what they teach, and I have specifically declined to resolve the ambiguous identifier "WO201114812" by assumption.
   • R1's filing/priority date and the '537's effective filing date (2012-10-23 vs. 2013-10-22) are unresolved; both are outcome-determinative.

Sources: https://patents.google.com/patent/US10009537/en · https://patents.justia.com/patent/10009537 · https://patents.google.com/patent/US20130155474A1/en · https://patents.google.com/patent/[US7362354B2](/patent/US7362354B2)/en · https://patents.justia.com/patent/8040382 · https://ptacts.uspto.gov/ptacts/ (SnapAid PTAB papers quoting '537 at 22:8–33) · https://npe.law.stanford.edu/patent/10009537

I'll research the prior art references cited in this patent's Background section and the family's PTAB record to ground the § 103 analysis.

I have the actual claim set emerging now. Let me pull the full claims and the specific prior art of record.

Obviousness Analysis — U.S. Patent No. 10,009,537 under 35 U.S.C. § 103

Date of analysis: April 26, 2026 · All patent/publication numbers are reproduced literally as they appear in the sources; none have been normalized or corrected.

---

0. Preliminary flags (read before relying on the analysis)

Flag 1 — Claim text now recovered; it contradicts the earlier-generated Litigation section.
The earlier "Litigation summary" section described the '537 as being asserted through a system claim ("a message provider operative, while said first logical criterion is still not satisfied… automatic image capturing only after a first logical criterion…"). That language belongs to the family's original/PCT-era claim set, not to the granted '537. The granted '537 claim 1 is a method claim, and I recovered it from two independent sources that corroborate each other:

• Justia's reproduction of the '537 claims (patents.justia.com/patent/10009537), and
• SnapAid's own PTAB papers, which quote "'537 Patent at 22:8–33" and reproduce the claim verbatim (petition 1558468, art. NfNR_shF47ohLzZQoiDttm2… and xW3IgwYYfyHzzEyvsK_gITg3fUQ…).

I treat the method claim set below as authoritative and flag the earlier description as in error for the '537 (it may describe the '348/'226/'682 claim sets). This does not affect the prior-art analysis, but it changes which claim is being attacked.

Flag 2 — Priority/effective filing date is unsettled and the analysis is date-sensitive. The '537 is a continuation carrying an effective filing date of either Oct. 22, 2013 (PCT/IL2013/050856) or the earlier provisionals (61/717,216, 61/759,643). Google Patents lists priority as 2012-10-23 while the specification says 61/717,216 was filed "Oct. 23, 2013." US 2013/0155474 A1 — the single most on-point reference — published June 20, 2013. Whether it is § 102/§ 103 prior art therefore turns on the priority date, and I could not confirm its filing date. I treat this as the principal unresolved vulnerability below.

Flag 3 — References I could not independently verify. The '537 front page's own "References Cited" list was not in the retrieved authoritative text. I could not verify the content of US 7,688,379; US 2009/0263028; US 2006/0017820; US 2006/0056835; US 8,125,557; "WO 201114812"; US 7,920,727; US 8,040,382 (partially verified, see §2) beyond the applicant's own one-line characterization. Where I rely on those, I say so and attribute the characterization to the applicant (which is itself an admission).

---

1. Framework

1.1 The claim to be tested — claim 1 of US 10,009,537 (verbatim, as recovered)

"1. A method for estimating quality of a digital image frame having pixels, where image is taken from a video stream comprising multiple images, each image frame comprises an image of the scene; where video stream was taken by a digital camera;

• obtaining a first value (QI1) associated with the digital camera at the time of said image capture;
• estimating a first weight (c1) associated with the first value;
• analyzing the captured image for detecting or recognizing one or more objects in, or one or more characteristics of, the image;
• obtaining a second value (QI2) associated with the analysis;
• estimating a second weight (c2) associated with the third value;
• calculating a total value according to, or based on, the first value (QI1) weighted according to first weight (c1), the second value (QI2) weighted according to second weight (c2);
• comparing the total value to a threshold;
• determining whether the total value is higher than, or below than, the threshold; and
• acting in response to the determining;
• where second weight (c2) is partially based respective values of previous images in said video."

Note the drafting defect: claim 1 introduces "a second value (QI2)" but then recites "estimating a second weight (c2) associated with the third value" — there is no third value in claim 1. A POSITA would read this as "associated with the second value (QI2)," and the '537 specification (e.g., the QI_ForTotal / w(t,C_i)_i formulas) confirms it. This is a § 112(b) concern, but for § 103 it is dispositive only of what must be shown: c2 is the weight on QI2. (Claim 10, a second independent claim, is drafted the same way with the same defect.)

Independent claim 10 adds: a single-enclosure device with a video digital camera module (lens + image sensor), a motion sensor consisting of or comprising an accelerometer, a gyroscope or both, and a processor; QI1 is associated with camera movement at the time of capture, "where movements were measured by inertial measurement unit attached or part of said camera."

Dependents: 2 (acting = affecting the digital camera), 3 (sound/pre-recorded oral message), 4 (controlling auto-focus based on QI2 or the total value), 5 (alerting a user), 6 (marking/transmitting/saving the image), 7 (threshold changes over image sequence based on total value), 8–9 (estimating c2 based on estimated error in the image analysis); 11–20 depend from claim 10 (16 saving, 17 adaptive threshold, 19, and 20 = motion/location sensor is an accelerometer, a gyroscope, a GPS, or a step counter).

1.2 POSITA

A person having ordinary skill in the art as of the effective filing date would have a B.S. in EE/CS (or equivalent) and ~2–4 years of experience in digital-camera image processing / computational photography, with working familiarity with: sharpness/blur metrics (DCT/frequency-domain, PSNR), exposure histogram thresholds, face detection, weighted metric combination, auto-capture on a quality criterion, and mobile sensor fusion (accelerometer/gyro). This is a low-to-moderate level of skill; the claim is a combination of known measurement and known arithmetic.

1.3 The KSR/Graham case

The '537 is a combination claim: known QI computation + known per-QI weighting + known total-vs-threshold comparison + known responsive action. The only candidate "inventive" hook is the final limitation — c2 partially based on values from previous frames. Crucially, the applicant admitted in the specification that weighting and prior-frame data are old: "Some implementations use a weight function in computation, but the weight is either constant or can be change by the user manually," and the specification's own confidence-level mathematics is built on "previous data," a formula the applicant presents as a general case, not a new one.

---

2. The prior-art record ("Prior Art" section of the '537, plus family PTAB art)

Record-level facts bearing on strength of any § 103 ground:

• Samsung charted claim 1 of the '537 in its E.D. Tex. invalidity contentions (per the earlier Litigation section) — so contentions exist, but I do not have the '537 chart.
• The five Samsung IPR/PGR petitions (IPR2025-01519/01520/01521/01522, PGR2025-00083) did not include the '537; the '537 was one of the three "unchallenged" patents. Institution was denied on discretionary grounds (Jan. 9, 2026) — not on the merits. A discretionary denial carries no probative weight for or against patentability, and SnapAid's § 101/§ 112 pleading points are not § 103 evidence.

---

3. Ground 1 — Claim 1 obvious over US 2013/0155474 in view of US 2009/0278958

This is the strongest ground because every element except the last is in the primary reference, and the last is supplied by a reference the applicant itself listed.

Motivation to combine ('474 + '958)

1. Same field and same problem. Both address in-camera, pre-/post-capture quality scoring of a frame within a frame sequence. "'958" is not merely analogous art — it is art the applicant itself cited as background, i.e., a POSITA is presumed to have it in hand.
2. Stated, articulable benefit. '958 gives the reason on its face: reducing redundant computation ("redundant calculation may be avoided") — a recognized engineering desideratum for the resource-constrained mobile devices '474 targets. Applying a prior frame's score to temper a current frame's weight is the "known technique to improve similar devices in the same way" rationale.
3. Temporal-indexing is inherent in '474. Because '474 evaluates a preview stream, its own scores are naturally indexed by frame; feeding a prior frame's score into the current weight requires no architectural change and yields predictable results (a smoother, less flickery quality score — the very stability the '537 says it wants).
4. No teaching away. '958 is a use-teaching, not a warning; nothing in '474, '958, '354 or '761 disavows using prior frames.
5. Reasonable expectation of success. Both references are computational and operate on the same data (frame scores); the combination is a matter of arithmetic.

---

4. Ground 2 — Claim 1 obvious over WO 2006/040761 (US 2007/0195174) in view of US 7,362,354 and US 2009/0278958

Use this if a tribunal gives '474's priority-date problem traction.

• '761 gives QI weighting and total grade expressly: "setting the captured image attributes relative weight for the compution of the total image grade."
• '761 gives the second QI as an object/characteristic QI, not merely a camera QI: scene dynamics include "motion speed of the subjects" — i.e., analysis of the content of the image, satisfying claim 1's "analyzing the captured image for detecting or recognizing… characteristics of the image."
• '354 (Lin) supplies the object-recognition QI with a concrete implementation (face quality figure of merit: detection, brightness/noise/contrast figures of merit, red-eye), each independently compared to a threshold, with feedback — and supplies claim 1's "detecting or recognizing one or more objects."
• '958 supplies the previous-frames weighting.

Motivation: all three are in in-camera image-quality assessment; '354's stated purpose ("provide appropriate photo quality feedback to the camera user so that the user can make an objective determination whether to keep or re-take the image") is the same purpose as '761's "total image grade" and the '537's stated aim. Combining a camera-state QI (from '761/'474) with a content/face QI ('354) is the obvious design choice of covering both "did the camera ruin it" and "is the subject good," and weighting them is a matter of design choice requiring no new hardware. KSR: "a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions" — here, additive weighted scoring is exactly that.

---

5. Ground 3 — Claims 10 and 20 obvious over US 8,040,382 in view of US 2013/0155474 (and '958)

Motivation: '382 and '474 attack the same defect — hand-shake blur — from opposite ends (predict-and-capture-at-minimum-motion vs. measure-and-inhibit-until-threshold). A POSITA seeking reliable, blur-free capture would combine them because they are complementary and non-redundant, with predictable results; '474 itself instructs the user to "hold the camera steadier," and '382 provides the sensor-based quantification of exactly that. Claim 20's laundry list of IMU/GPS/step-counter is a routine recitation of alternative known sensors doing what they are known to do (KSR; In re KSR "a mere duplication of parts" / aggregation).

---

6. Dependent claims

Note the important negative: claim 13-type limitations and the "gesture-based shutter/delete," "vehicle speed" and "video quality" subject matter discussed in the earlier Litigation/Patent-summary sections appear in the Embodiment list of the specification, but I could not verify whether they are recited in any granted '537 claim. If they are only in the specification, they are irrelevant to § 103 of the claims.

---

7. Anticipation vs. obviousness — why § 103 is the right theory

I looked for a clean § 102 reference and none anticipates claim 1:

• '474 discloses every element except an express weighting of QI1/QI2 and the previous-frames basis for c2. Its "group threshold or total overall quality score" is disjunctive and does not necessarily weight.
• '761 discloses weighting and a subject-motion QI and a total grade, but I have no verified disclosure of the previous-frames weighting limitation.
• Anon discloses a "weighted, combination of parameters" and suggestions, but (as Samsung's own framing shows) via rule sets/lookup data, not previous-frame scores.
• Takeuchi discloses a "total evaluation result" from a "weighted combination of different quality evaluation methods" including explicit over/under-exposure quantification — again, not framed on previous frames.

So the defensible theory is obviousness of a combination, with the c2-from-previous-frames limitation carried by US 2009/0278958 (express) and/or US 8,040,382 (frame-adjacency) — both of which the applicant put on the record itself.

---

8. Counterarguments the owner will make, and how they fare

1. "'958 is cited only for scoring a base image, not for weighting a current frame's QI." Weak: '958's stated rationale ("redundant calculation may be avoided") is expressly a cross-frame computation-sharing rationale, which is what the '537 does.
2. "The '537's insight is interdependency — one QI's weight depends on other QIs' values, weights and confidence levels." This argument is available for the family's broader claims (the '901/'325/'682 claim sets recite confidence levels and cross-QI dependency) but is materially weaker for '537 claim 1, which requires only that c2 be "partially based respective values of previous images," not on other current QIs. Owner is likely to conflate the '537 with its siblings — a distinction worth pressing.
3. "The Examiner allowed for the specific combination." SnapAid's own PTAB papers assert "the prior art does not teach analyzing the captured image for detecting or recognizing one or more objects… [in combination with] second weight (c2)… based on previous images." That is a prosecution-history argument, not evidence; and the record shows the cited art was never combined against '474 + '958.
4. Non-obviousness by unexpected results / long-felt need. The '537 specification supplies no comparative data and no evidence of failure of others; "users take 5–6 pictures" is a general complaint, not a long-felt need tied to the claimed combination. Under Graham, this prong is essentially unpopulated.
5. Priority-date attack on '474 (Flag 2) — the owner's best procedural weapon. If '474's filing/priority date is confirmed to be on or after the '537's effective filing date, '474 drops out as § 102(a)(2)/§ 103 art and Ground 1 must be rebuilt. Ground 2 ('761 + '354 + '958) is the fallback, and '761 ('04/'07) and '354 ('02/'03) and '958 ('09) are all safely earlier on any priority theory.

---

9. Evidence that would still be needed to make these grounds institution-grade

1. The printed claim set of the '537 as granted (I reconstructed claim 1 from two corroborating secondary sources; a certified copy should be used). Confirm whether there are exactly two independent claims and that claim 10's "third value" defect is as-recovered.
2. '474's filing and priority dates — outcome-determinative for Ground 1.
3. A POSITA declaration mapping each element of Grounds 1–3, with particular attention to the "estimating a first weight" element in '474 (does "combined into a group threshold" teach weighting, or does '761 need to be added?).
4. The '537 prosecution file to see exactly which claims/references were before the Examiner — SnapAid asserts the prior art "did not teach" the combination, which implies the Examiner did not have '474 + '958 before him in combination.
5. Secondary-consideration evidence (none presently in the record).

---

10. Confidence summary

Explicitly unresolved: '474's dating; the full "References Cited" list of the '537; whether US 7,688,379 / US 2009/0263028 / US 2006/0017820 / US 2006/0056835 / US 8,125,557 / "WO 201114812" / US 7,920,727 supply additional elements (I did not verify their contents and do not assert what they teach beyond the applicant's one-line characterizations).

Sources: https://patents.google.com/patent/US10009537/en · https://patents.justia.com/patent/10009537 · https://patents.google.com/patent/US20130155474A1/en · https://patents.google.com/patent/[US7362354B2](/patent/US7362354B2)/en · https://patents.justia.com/patent/[8040382](/patent/8040382) · http://data.epo.org/pise-server/rest/collections/lgpi/EP1474779A2.pdf · https://ptacts.uspto.gov/ptacts/public-informations/petitions/1558468 (Samsung petition, '325/'901 grounds citing Anon & Takeuchi; SnapAid § 101 response quoting '537 at 22:8–33) · https://ptacts.uspto.gov/ptacts/public-informations/petitions/1558467 · https://www.freepatentsonline.com/y2015/0288874.html · https://www.patents-review.com/a/20170237900-real-time-assessment-picture-quality.html