Neuromorphic Vision Sensor Patents: Leaders, Trends & Gaps 2026
Filing growth compares 2021 (4 records) with 2024 (2) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field.
What this landscape covers
This review tracks 30 published patent records filed between 2015 and mid-2026 that describe neuromorphic vision sensors and adjacent neuromorphic computing hardware — event-driven imaging elements, sensor-processor fusion architectures, and the pictorial-communication and image-recognition circuitry that surrounds them. The scope is narrow by design: it isolates sensor-side neuromorphic claims from the much larger neuromorphic-processor literature, so the picture here is of a specialised, still-forming niche rather than a mature computing category.
Coverage runs through a 2026-07-31 data cut-off. Because publication typically lags filing by around 18 months, the 2025 and 2026 figures are undercounts and should be read as provisional rather than as evidence of a slowdown.
Let an AI agent run this analysis on your own technology
Pick a task. Every answer cites the patents behind it.
Filing trend and technology composition
Two views of the same 30-record dataset: how filing volume has moved year over year, and which IPC subclasses carry the claim density within it.
Filing trend, 2017–2026
Filings rose from 3 in 2017 to a peak of 9 in 2022, then declined; the 2021-to-2024 window shows a documented drop from 4 to 2 records, a 50% fall. Treat 2025 and 2026 as incomplete given the typical 18-month publication lag rather than as a continuation of that decline.
IPC subclass composition
H04N (pictorial communication) appears in 40.0% of the 30 records, the largest single class, followed by G06V (image/video recognition) at 23.3% and G06T (image data processing) at 20.0%. Because records can carry multiple classes, these shares sum to more than 100%; G06N (AI-model computing) and G06F (digital data processing) sit at the lower end, at 10.0% and 6.7% respectively.
Shares are the percentage of the 30 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Neuromorphic Computing: Neuromorphic Vision Sensor Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about neuromorphic computing: neuromorphic vision sensor patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative and most-cited filings
WO2026054852A2 — AI-based reconfigurable neuromorphic vision sensor fusion systems
The filing describes an AI-based, reconfigurable neuromorphic vision sensor fusion platform built from neuromorphic functional sensing units arranged in flat or dome-shaped arrays, paired with an underlying processing stack of AI processing layers organised in a top-to-bottom bio-morphic hierarchy. Its processing modules and interfaces can couple and uncouple from neighbouring modules in real time during operation.Filed by Srivastava, Deepak; published 2026-03-12.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20180173992A1 | Vector engine and methodologies using digital neuromorphic (NM) data | 18 |
| 2 | US10229341B2 | Vector engine and methodologies using digital neuromorphic (NM) data | 15 |
| 3 | US20220377222A1 | System with adaptive light source and neuromorphic vision sensor | 10 |
| 4 | US10789495B2 | System and method for 1D root association providing sparsity guarantee in image data | 8 |
| 5 | CN117893578A | 基于互补神经形态视觉的光流场计算系统 | 5 |
| 6 | WO2025123803A1 | 基于互补神经形态视觉的光流场计算系统 | 3 |
| 7 | WO2023078387A1 | System and methods for ultrafast widefield quantum sensing using neuromorphic vision sensors | 2 |
| 8 | US20230238406A1 | Bio-Inspired Imaging Device with In-Sensor Visual Adaptation | 2 |
| 9 | US20240292074A1 | Neuromorphic compressive sensing in low-light environment | 2 |
| 10 | US20230050794A1 | Cone-rod dual-modality neuromorphic vision sensor | 2 |
Citation counts are drawn from within this searched corpus and favour older filings; treat them as a signal of influence on the field to date, not of current commercial relevance.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the numbers mean for a filing decision
Three findings stand out once the ranking, the trend and the IPC composition are read together.
A leader, not a monopoly
The top-ranked assignee holds 6 of the records in the ranking, with the field dropping to 4 records by fifth place and just 1 by tenth. That shape — one clear leader over a long tail of single- and dual-filing entrants — means the space is still open to a well-drafted claim rather than locked up by one player.
Peak has passed, but read the tail carefully
Filing volume peaked at 9 records in 2022 and had fallen to 2 by 2024, a documented -50% move from the 4 filed in 2021. That is a real contraction over a complete three-year window; 2025 and 2026 figures are too fresh, given typical publication lag, to extend that story further.
Imaging and recognition carry the density
Pictorial communication (H04N) and image/video recognition (G06V) between them cover the largest shares of the 30 records in scope, showing where drafting has concentrated. AI-model computing (G06N) and core digital processing (G06F) sit far lower, at 10.0% and 6.7%, suggesting the sensor-fusion and learning-architecture layers are less crowded than the imaging front end.
One automotive cluster files together
The strongest co-assignee links in the dataset run between three automotive manufacturers, each pairing appearing on 4 records. That is a deliberate joint-filing pattern rather than coincidence, and it marks a cluster worth watching for coordinated claim strategy around vehicle-mounted neuromorphic sensing.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to neuromorphic computing: neuromorphic vision sensor patent landscape, with the prior art for and against each one.
| Assignee | Co-assignee | Shared families |
|---|---|---|
| Audi AG | Volkswagen AG | 4 |
| Audi AG | Dr. Ing. h.c. F. Porsche AG | 4 |
| Volkswagen AG | Dr. Ing. h.c. F. Porsche AG | 4 |
Three co-assignee pairs recur in the dataset, all linking the same trio of automotive manufacturers at 4 shared records each — a signal of a joint development programme rather than incidental overlap.
Where to take this analysis
The dataset points to specific next steps depending on whether the goal is freedom-to-operate, portfolio building, or scouting for licensing partners.
Check freedom-to-operate against the top filer
With one assignee holding 6 records against a field of 13 ranked companies, any new filing in sensor-fusion architecture should be checked against that leader's claim scope before drafting begins.
Run a freedom-to-operate search in EurekaWatch the automotive co-filing cluster
Three automotive manufacturers co-file consistently at 4 shared records per pairing — a pattern worth monitoring for signs of a shared roadmap in vehicle-mounted neuromorphic sensing.
Track this cluster in EurekaProbe the AI-model and digital-processing classes
G06N and G06F carry the lowest shares of the 30 records in scope, which may indicate room for claims that tie sensor output directly to on-chip learning architectures.
Explore white space in EurekaCommon questions on neuromorphic vision sensor patents
The ranked leader in this dataset holds 6 records, ahead of a field that drops to 4 records by fifth place and just 1 by tenth. The ranking covers 13 companies in total, so this is a leader-plus-long-tail pattern rather than a market dominated by one or two firms. A newcomer's freedom-to-operate check should start with that leader's claim scope, but the long tail means there is meaningful room for differentiated filings elsewhere in the space.
Filing volume peaked at 9 records in 2022 and had fallen to 2 by 2024, a documented -50% change from the 4 filed in 2021 — a real contraction over a complete window. However, publication typically lags filing by around 18 months, so the 2025 and 2026 counts in the raw data are undercounts, not evidence that filing has stopped. The honest read is a post-peak cooling through 2024, with the most recent years still filling in.
Within the 30 records in scope, pictorial communication (H04N) appears most often at 40.0%, followed by image/video recognition (G06V) at 23.3% and image data processing (G06T) at 20.0%. Material analysis (G01N) and AI-model computing (G06N) each sit at 10.0%, and core digital processing (G06F) and laser/stimulated-emission hardware (H01S) each cover 6.7%. Because a single record can carry several IPC classes, these figures do not sum to 100% and should be read as independent shares of the record total, not as a breakdown of the whole.
The dataset records three co-assignee pairs, and the strongest links all involve the same trio of automotive manufacturers, each pairing appearing on 4 shared records. That consistency across all three possible pairings among the same three companies points to a coordinated filing programme rather than incidental overlap, likely tied to vehicle-mounted sensing development. Anyone assessing competitive risk in automotive neuromorphic vision should treat this cluster as a single strategic unit rather than three independent filers.
The IPC composition shows AI-model computing (G06N) and core digital data processing (G06F) carrying the lowest shares of the 30 records, at 10.0% and 6.7% respectively, compared with 40.0% for pictorial communication claims. That gap suggests the imaging and signal front end is more heavily claimed than the learning and processing architecture that sits behind it. A first claim tying sensor-level event data directly to an on-chip adaptive learning module would sit in a less crowded part of this landscape than another imaging-circuit claim.
Research Neuromorphic Computing: Neuromorphic Vision Sensor Patent Landscape in depth with Eureka
Go past this page: query the whole neuromorphic computing: neuromorphic vision sensor patent landscape corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.