1
\$\begingroup\$

I found in many Android devices that the "real array size", which is the actual pixel count of the camera sensor, is lower than the 4:3 Max picture the user can take. For example: Sensor Array = 1616 * 1232 = 1.99 MP, and Camera Application can take a 5 MP Picture.

Can we say that in this case the sensor is "extrapolated"? or "boosted"? or how to name this when the picture taken is different in max megapixels taken with sensor array size?

\$\endgroup\$
0

2 Answers 2

4
\$\begingroup\$

Yes, if the sensor is truly a 1600x1200 (two megapixel) device and the camera is artificially generating 5 megapixel images, then yes, one could say "extrapolated" — or, dishonest, pointless scammery, since this doesn't provide you with any real advantages.

To be pedantic, we wouldn't say that the sensor is extrapolated. We would say that the 5mpix image is extrapolated from the actual data from the smaller sensor.

It's more correct to say interpolated rather than extrapolated — and indeed you will often find this term buried in specifications or in reviews. That's because the made-up data is constructed from the surrounding points, rather than being projected outside of it. But at this scale, it doesn't really matter, since it's so far from the range where added values are providing any actual useful information.

I hope that by now you won't see this on any but the very cheapest knockoff phones, because genuine reasonable-quality 5mpix and 8mpix parts are readily available at very low cost. But, with a quick search, I found an example, in a $100 unlocked 3G phone — although they're at least honest enough to list the real number first, rather than in the fine print.

\$\endgroup\$
4
  • \$\begingroup\$ there the problem is: how can someone know the real sensor vendor by making a tear down regarding the technical data sheets? \$\endgroup\$
    – M. A.
    Jun 20, 2015 at 19:47
  • 1
    \$\begingroup\$ I'm pretty sure that there is no extrapolation going on. The additional pixels are more likely added in between existing ones, hence interpolation seems to be the more appropriate term. \$\endgroup\$
    – null
    Jun 20, 2015 at 19:50
  • \$\begingroup\$ @null You're right. Edited. \$\endgroup\$
    – mattdm
    Jun 20, 2015 at 19:58
  • \$\begingroup\$ @null Well noted! \$\endgroup\$
    – M. A.
    Jun 20, 2015 at 20:09
1
\$\begingroup\$

A few cameras that have image stabilization built into the body can take a sequence of images with the sensor moved slightly in each one. The images are then combined to generate one image with higher resolution than the sensor can provide in a single frame. It's possible that phone-based cameras might do something similar using an accelerometer to detect device movement. I don't have a name for the technique, but you could say the image is a composite image.

\$\endgroup\$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.