Machine Vision Patents: Top Companies & Filing Trends 2026
- 62.7% of all 161 records sit with just five assignees, and 79.5% with ten — this field is unusually top-heavy for its size.
- Filing peaked in 2021 at 30 records then fell to 14 by 2024, a -53% drop over that three-year span, before the still-filling 2025-26 tail.
- Image processing and control classes dominate G06T and G06F together cover over 40% of records, while conveying and semiconductor-handling classes point to where vision meets physical automation.
Filing growth compares 2021 (30 records) with 2024 (14) — 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. Top-5 share is the combined record count of the five largest assignees divided by all 161 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This dataset tracks 161 published records filed between 2015 and the August 2026 data cut-off, drawn from filings that combine visual inspection or vision-guided automation with process-control concepts such as fault diagnosis, equipment state monitoring or failure prediction. That combination narrows the field to machine vision as it is deployed inside industrial control loops, rather than vision research in general.
The receiving-office spread is heavily weighted toward the United States, with Europe, WIPO/PCT and India as the next largest venues. Publication lags filing by roughly 18 months, so the most recent one to two years in any trend understate actual filing activity.
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 161 records: how filing activity has moved year over year, and which IPC subclasses carry the claim density.
Filing trend, 2017-2026
Filings rose from 18 in 2017 to a peak of 30 in 2021, then declined to 14 by 2024 — a -53% move over that three-year span. 2025 and 2026 figures are still filling in under the usual publication lag and should not be read as a continued decline.
Technology composition by IPC subclass
G06T (image data processing) leads at 23.0% of the 161 records, followed by G06F (digital data processing) at 20.5% and G05B (control and regulating systems) at 16.8%. H01L, B65G, G06V, H05K and H10P each sit near or above 9%, showing that vision claims here are frequently tied to semiconductor handling and material-conveying equipment rather than pure image algorithms. Records can carry multiple classes, so these shares sum to more than 100%.
Shares are the percentage of the 161 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Machine Vision Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about machine vision patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative and most-cited filings
US11764097B2 — System and method for ring frame cleaning and inspection
A system and method for cleaning and inspecting ring frames is disclosed here. In one embodiment, a ring frame processing system includes: a cleaning station configured to remove a first tape on a first surface of a ring frame using a first blade, clean first adhesive residues from the first tape on the first surface of the ring frame using a first wheel brush, and remove second adhesive residues from a second tape on a second surface of the ring frame using a second blade; and an inspection station, wherein the inspection station comprises an automated optical inspection system configured to determine the cleanness of the first and second surfaces of the ring frame after cleaning.Filed by Taiwan Semiconductor Manufacturing Company, granted 2023-09-19.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20170080255A1 | Thin and wearable ultrasound phased array devices | 121 |
| 2 | US20040004482A1 | Industrial inspection using combination of functional testing and structural inspection | 61 |
| 3 | US20230125477A1 | Defect detection using one or more neural networks | 57 |
| 4 | WO2021062536A1 | System and method for ai visual inspection | 44 |
| 5 | US20220366558A1 | System and method for ai visual inspection | 37 |
| 6 | US20190047799A1 | Handling system with independent and coordinated shuttle, for industrial automation | 37 |
| 7 | US20220147026A1 | Method and Apparatus for Improved Auto-Calibration of a Robotic Cell | 36 |
| 8 | US20100122633A1 | Vertically separated pass through conveyor system and method in surface mount technology process equipment | 36 |
| 9 | US8413578B2 | Modular printing system having vertically separated pass through conveyor system | 35 |
| 10 | US20210142456A1 | Image Analysis System for Testing in Manufacturing | 28 |
Citation counts favour older filings and signal influence within this searched corpus, not current commercial importance.
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 data shows
Three patterns stand out once filing volume, technology class and citation weight are read together.
A small group holds most of the claim space
Five assignees account for 101 of the 161 records in scope, and ten account for 128. That is a tighter concentration than the raw filer count of 35 ranked companies would suggest, and it means most white space sits outside the leading names rather than between them.
Peak filing has passed, but the picture is incomplete
Filing activity peaked in 2021 at 30 records and fell to 14 by 2024, a -53% move. Because publication lags filing by about 18 months, 2025 and 2026 counts are not yet complete and should not be read as confirming a continued fall.
Image processing leads, but control and hardware classes are close behind
G06T (image data processing) is the largest single class at 23.0% of records, with G06F and G05B close behind. The presence of H01L, B65G and H05K at meaningful shares shows that a large share of this field's claims are anchored to specific hardware contexts — semiconductor handling, conveying, circuit assembly — rather than to vision algorithms alone.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to machine vision patent landscape, with the prior art for and against each one.
The competitive set
The ranking covers 35 companies with any presence in this dataset — not a top-50 or top-100 cut — and activity is heavily skewed toward a handful of names.
One filer sits well ahead of the field
The leading assignee holds 31 records, close to a fifth of all 161 in scope on its own. The gap to fifth place, at 11 records, is wide, and by tenth place volume has thinned to 3 records — a classic head-and-long-tail shape.
Most ranked filers hold only a handful of records
Beyond the leading names, the ranking thins out quickly. Companies entering with a single filing are common in this dataset, which typically signals either a specific product application or an early-stage entrant testing the space rather than a sustained filing programme.
Several top filers show no activity in the latest tracked year
Multiple leading assignees, including the top-ranked filer, show zero records in the most recent year, with some down -100% year over year. Given the roughly 18-month publication lag, this is consistent with filings still working through the pipeline rather than firms exiting the space.
| Assignee | Recent year | YoY |
|---|---|---|
| Dell Products L.P. | 0 | -100% |
| Illinois Tool Works Inc. | 0 | — |
| Bright Machines Inc. | 0 | -100% |
| Vismunda | 0 | — |
| Rockwell Automation Technologies, Inc. | 0 | -100% |
| Robert Bosch GmbH | 0 | -100% |
| Taiwan Semiconductor Manufacturing Co., Ltd. | 0 | — |
| MUSASHI AUTO PARTS CANADA INC | 0 | — |
Where to take this analysis
The dataset points to specific next steps depending on whether the goal is defensive clearance or offensive filing strategy.
Map claim scope against the leading assignees
With 62.7% of records held by five companies, understanding exactly what those claims cover — not just how many they hold — is the first filter before filing near this space.
Explore assignee claim scope in EurekaWatch the 2024-2026 filing tail as it fills in
The apparent decline after 2021 is partly a publication-lag artefact. Tracking how 2025 filings revise upward over the next reporting cycles will clarify whether the field is genuinely cooling.
Set up filing trend tracking in EurekaScan the under-claimed branches before drafting
Sub-areas like vision-guided fault diagnosis on control networks show thinner density than the core image-processing classes, making them worth a closer prior-art pass before committing claim language.
Run a white space scan in EurekaCommon questions on this landscape
The assignee ranking in this dataset covers 35 companies, and it is heavily concentrated: the top five together hold 62.7% of the 161 records in scope, and the top ten hold 79.5%. The leading assignee alone accounts for 31 records, well ahead of fifth place at 11. This means a small number of firms occupy most of the claim space, while the remaining ranked companies each hold only a handful of filings, often tied to a specific application.
Filing activity peaked in 2021 at 30 records and had fallen to 14 by 2024, a -53% move over that three-year span. However, publication typically lags filing by around 18 months, so 2025 and 2026 figures in any dataset are still incomplete and should not be read as evidence of a continuing decline. The safest reading is that 2021 was a high point and 2024 is the most recent year that can be treated as a complete count.
Image data processing (IPC class G06T) is the largest single category at 23.0% of the 161 records, followed by general digital data processing (G06F) at 20.5% and control and regulating systems (G05B) at 16.8%. Semiconductor devices, conveying and material handling, image recognition, and printed-circuit assembly classes each sit near or above 9%, showing that vision claims in this field are frequently tied to a specific hardware or process context rather than filed as standalone algorithms.
US11764097B2, assigned to Taiwan Semiconductor Manufacturing Company, claims a ring frame processing system that combines a cleaning station with an automated optical inspection station to verify surface cleanliness after tape and adhesive removal. It is a specific hardware-plus-inspection combination tied to semiconductor ring frame handling, not a general vision-inspection method. Work that uses automated optical inspection for different substrates, different cleaning mechanisms, or without the specific cleaning-then-inspect sequence described is less likely to fall inside its claim scope, but a freedom-to-operate review of the granted claim language is still advisable before designing directly around it.
Relative to the dense core classes of image processing and control systems, sub-areas such as vision-guided fault diagnosis on control networks, multi-camera equipment-state fusion for predictive maintenance, and automated optical inspection applied to flexible material handling show comparatively thinner filing density. These are not guaranteed to be open — a full prior-art search is still required — but they are reasonable starting points for a novelty scan given the composition of this 161-record dataset.
Research Machine Vision Patent Landscape in depth with Eureka
Go past this page: query the whole machine vision 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.