Global Shutter in Machine Vision

A global shutter exposes every pixel on a camera sensor at exactly the same moment. This simultaneous image acquisition method ensures that the entire frame represents a single, uniform snapshot in time. Unlike architectures that read out line-by-line, this approach eliminates spatial distortion, making it indispensable for applications where objects are moving rapidly when the camera fires. On a fast conveyor belt or inside a robotic workcell, the geometry of the image must be completely true to reality. This architecture guarantees it.

How a global shutter works

The key to this technology is what happens inside the individual pixel. In a global shutter CMOS sensor, a small analog memory node, a shielded holding area, sits between the photodiode and the readout circuitry.

When the camera triggers an exposure, every photodiode on the sensor collects light simultaneously. Once the precise exposure time elapses, every pixel instantly transfers its accumulated charge into its dedicated memory node. The sensor can then read out the data row by row and transmit it via standard interfaces like GigE Vision or USB3 Vision at its own pace. Because the charge is shielded in the memory node during readout, further light hitting the sensor does not affect the image. The moment is already locked in.

This capability introduces a physical tradeoff. The analog memory node takes up space on the silicon, leaving less room for the light-gathering photodiode. This reduced fill factor generally means a global shutter sensor will have slightly lower quantum efficiency than a rolling shutter sensor at the identical pixel pitch.

Where Global Shutter matters

In industrial automation, these sensors are the default choice when objects are moving. When a vision system must inspect, measure, or track items at high speed, engineers rely on this architecture to ensure absolute geometric accuracy.

Sensors engineered specifically for these rigorous industrial demands, such as the Sony's Pregius and Pregius S series, have advanced the architecture significantly. By redesigning the pixel structure, these modern sensors maximize dynamic range and minimize noise, closing most of the historical sensitivity gap between global and rolling designs.

Global shutter vs. rolling shutter: which do you need?

The choice between these two architectures comes down to whether motion during image capture will distort the image enough to hurt the inspection, measurement, or decision-making task.

If a fast-moving object passes in front of a rolling shutter sensor, the top of the object is recorded slightly earlier than the bottom. This time delay causes straight lines to lean and circular objects to appear oval. If your software requires exact sub-pixel measurements of a moving component to pass or fail it, this distortion will break your inspection logic. A global shutter is mandatory here.

Conversely, if the object is stationary, or if you are inspecting moving objects inside a dark enclosure using a microsecond strobe light to artificially freeze the scene, a global shutter is unnecessary. In those specific scenarios, a rolling shutter provides better light sensitivity and higher resolution for a lower cost.

Key specifications to consider

When integrating a camera with a global shutter, evaluate these metrics to match the hardware to your system's demands: