Intelligent Lighting - Part 1

HOW SMART CAN MACHINE VISION

ILLUMINATION BE? >

Learn more about intelligent lighting | Part 1

Read more

Binning

Binning denominates a technique where the intensities of neighbouring pixels on a sensor are combined.

Binning is the term used to describe the technique of combining the output of adjacent pixels on a sensor. This can be performed on pixels in a single axis or, more commonly, in both directions. It will usually be shown as "2x horizontal binning" or "4x4 binning" where the numbers show the size of the cluster of pixels that is combined. This technique has a detrimental effect on the spatial resolution, but increases sensitivity, speeds up frame rates and increases the signal-to-noise ratio (SNR) .

The example before shows 2x2 binning on a CCD sensor. The spatial resolution is halved in each direction, but the sensitivity is increased by a factor of 4. Because the output of two pixels is combined, the readout time is halved, thus doubling the frame rate. With each readout event, associated noise is generated in the signal, so by reading out four pixels at a time this readout noise is quartered. The well capacity of the 'superpixel' is 4x that of a single pixel, so the dynamic range is subsequently higher.

This technique was traditionally only applied to monochrome or 3-chip cameras because using this with sensors with Bayer filters will result in wrong colours. However, some manufacturers have already patented methods of binning using Bayer sensors which involves holding the charge in multiple shift registers before summing the appropriate colours prior to being read out.