TB-6251 · REV E · Technical newsheet

Machine Vision & InspectionDevice profile

OMNIVISION Adds Two Global Shutter CMOS Sensors for Industrial Vision

OMNIVISION has announced two new global shutter CMOS image sensors for industrial machine vision, factory automation, ITS, and robotics, but disclosed no part numbers, resolution, frame rate, or pricing, leaving benchmark comparisons to incumbent families pending full datasheet r

By Sophie Lindqvist4 min read740 words

Features

  • OMNIVISION announced two new global shutter CMOS image sensors for industrial vision and robotics.
  • No part numbers, pixel counts, frame rates, or pricing were disclosed at the announcement stage.
  • Targeted application classes are industrial machine vision, factory automation, intelligent transportation systems, and robotics.
  • Competing global shutter families already in market include Sony's Pregius IMX line, onsemi's Python and XGS series, and Hamamatsu's Orca Flash sensors.
  • Shutter efficiency above 90 percent is the working minimum for quantitative machine vision inspection applications.

OMNIVISION has released two new global shutter CMOS image sensors targeting industrial machine vision, factory automation, intelligent transportation systems (ITS), and robotics, expanding its portfolio in a category defined by simultaneous pixel exposure rather than line-by-line scanning.

The announcement provides no part numbers, pixel counts, frame rates, or pricing. OMNIVISION has yet to publish full datasheets, so resolution class, optical format, and interface standard remain to be confirmed through later product documentation.

Why a global shutter matters in motion-critical work

Global shutter sensors expose every pixel in the array during the same instant, then transfer the collected charge to a shielded storage node for readout. Rolling shutter imagers scan rows sequentially, so a straight translation of an object across the frame bends into a shear because rows captured at different times register different object positions.

For conveyor tracking at industrial line speeds, pick-and-place robotics indexing workpieces on sub-millisecond cadence, license-plate recognition at highway velocity, and simultaneous localization and mapping (SLAM) on mobile robots, that distortion translates directly into position errors. Global shutters eliminate the shear at the cost of more complex pixel architectures—typically a pinned photodiode plus a dedicated storage gate per pixel—which historically reduced fill factor and quantum efficiency compared with rolling shutter devices. Modern back-side illuminated (BSI) global shutter pixels and stacked-die designs have largely closed that efficiency gap.

What an industrial buyer usually checks first

Vision integrators evaluate global shutter sensors on a defined metric set:

  • Resolution and pixel pitch set the spatial sampling budget against the smallest feature the application must resolve.
  • Frame rate sets the temporal sampling budget—90 fps covers many conveyor lines, while 200 fps and above serve robot guidance under vibration.
  • Shutter efficiency, also called global shutter efficiency or parasitic light sensitivity (PLS), quantifies unwanted light leakage into the storage node during readout; values above 90 percent are the working minimum for quantitative inspection.
  • Dynamic range, expressed in dB or as a full-well ratio, governs whether both specular highlights and shadow detail survive a single exposure.
  • Interface options—MIPI CSI-2 for embedded platforms, GigE Vision, USB3 Vision, CoaXPress, or 10 GigE for PC-based systems—dictate cable reach, bandwidth, and frame-rate headroom.

The release does not yet specify which of these parameters the two new parts target. Without a resolution figure, frame rate, optical format, or interface standard from the vendor, integrators cannot benchmark the sensors against incumbents such as Sony's Pregius IMX family, onsemi's Python and XGS lines, or Hamamatsu's Orca Flash series.

Application drivers behind the expansion

Four application classes shaped the product description OMNIVISION published. Machine vision in quality inspection demands repeatable geometry from frame to frame. Factory automation in Industry 4.0 cells requires sensor nodes small enough to mount in tight fixtures, often with Power over Ethernet (PoE) to simplify cabling. Intelligent transportation systems—toll gates, tunnel inspection, and traffic analytics—need sensors that hold exposure integrity under mixed artificial and natural lighting. Robotics, particularly autonomous mobile robots (AMRs), rely on global shutter data to feed visual odometry and obstacle detection pipelines whose accuracy compounds over distances of meters.

Each class imposes different constraints. Factory cells prefer compact modules with industrial-rated interfaces. AMRs prioritize low power and small form factor. ITS sites require high dynamic range to handle headlamp glare against dark pavement. None of these constraints appears in the announcement text, leaving the application-to-sensor mapping unresolved.

What the product release will need to answer

The next development the channel will watch for is a full datasheet drop. The announcement establishes market intent but defers the buying decision until concrete numbers appear—resolution class, frame rate, optical format, interface compliance (such as the MIPI CSI-2 revision level supported), operating temperature range, and conformance with any relevant industrial or automotive qualification standards.

Until those specifications surface, the release reads as a product placeholder rather than a comparison-ready part. For systems integrators, the open question is whether the two new sensors will land in cost-sensitive price tiers or aim for the high-performance segment where shutter efficiency, dynamic range, and embedded processing already compete among established vendors. Compliance targets—whether the parts are specced for industrial environments (such as extended -40 °C to +85 °C operating ranges) or specific automotive safety standards—will determine which of the four named application classes the devices can credibly enter.

via ovt.com (Original)

Filed under

  • global-shutter
  • cmos-image-sensor
  • machine-vision
  • omnivision
  • factory-automation
Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

News editor covering marketplaces and e-commerce at Testbench Report.

66 articles

Application notes

« Previous articleNext article »