What Is the Equivalent to ISO Values in Machine Vision Cameras#
About ISO Values#
The ISO value is a measure of a photographic film's sensitivity to light. It is also known as the "film speed" and is related to the size of the photosensitive grain on film plates.
The term comes from the analog photography era and is today still used also for digital cameras. However, we will find the term mostly in digital consumer photo cameras, such as DSLR or smartphone cameras. Here, the user can change the ISO value to compensate bad environment lighting conditions or just to snap brighter images. A higher ISO increases the sensitivity and leads to brighter images.
Common values are ISO 100 / 200 / 400 / 800 / 1600 / 2400 / 3200 / 6400, where ISO 100 is the basic sensitivity in most cases, or better - an amplification of 0. The next ISO value would double the light sensitivity (one "stop") which in turn means it also doubles the brightness of an image.
However, those values are not standardized and could mean different brightness conditions on different sensors and cameras, as every manufacturer defines the start values for himself.
ISO Values in Machine Vision Cameras#
Talking about machine vision cameras, there is no feature like ISO named in the SFNC (Standard Features Naming Convention).
The closest equivalent to film speed or ISO is the Gain camera feature. The gain basically can decrease or increase the image brightness by amplifying the signal received from the pixel cells of a sensor. However, the gain in a machine vision camera can be set in much finer steps than it's usually possible with ISO.
Most of the Basler cameras have the gain given in dB and within a 0–24 dB range. Jumping from 0 dB to 6 dB and every 6 dB further actually doubles the image brightness. The following table shows an example comparison between a photo camera ISO values and digital video camera gain values. The values below are only example values that have not been tested.
| ISO Value - DSLR Camera | Gain [dB] - Machine Vision Camera |
|---|---|
| 100 | 0 |
| 200 | 6 |
| 400 | 12 |
| 800 | 18 |
| 1600 | 24 |
Why Is There No ISO Feature in Machine Vision Cameras#
First, this has historical reasons. ISO was developed to standardize the film material used in still cameras or in cameras used for movie productions.
Gain instead is the default term for electronic signal processing. Cameras which always had an electronic sensor like cameras used for TV broadcasting or early video cameras using a videotape use an electronic signal. The term gain is used when such a signal (or any other electronic signal) is amplified.
A second explanation is the different expectation to the final image result. While still cameras often try to achieve the best picture using all sorts of image enhancement method's machine vision cameras usually want to provide a reproducible real image without any processing.
For that purpose, Basler differentiates between analog gain, digital gain, binning, or gamma settings to allow the best possible control and knowledge about the pixel processing. To work with only one linear ISO value would mean hiding and changing all these other parameters under the hood.
Conclusion#
There is no way to convert values from ISO to gain or vice versa, as there is no direct relation between both definitions. One describes the sensitivity of chemical film material, while the other the amplification of an electronic signal.