‘Modulo Camera’: researchers at the MIT are developing the HDR sensor


deutsch_s1

It is already the second time in the last month that scientists at the MIT have presented a new technology that might be able to make the life of every single photographer a lot easier. Less than two weeks ago a different team at the renowned university in Cambridge introduced an algorithm developed in cooperation with Google which can remove reflections and other annoying foreground elements from a photo. A few days ago the concept for a sensor with an unlimited Full-Well capacity was introduced. In simple terms, this is a sensor which makes overexposing an image impossible.

mit_modulo_camera_2

Image source: MIT


mit_modulo_camera_3

Image source: MIT

This is achieved by individually accessing the photodiodes on the sensor and resetting them if they are fully saturated (i.e. overexposed). This means that every “pixel” of the sensor can be exposed multiple times in the course of one exposure, with the image processor noting how many times individual “pixel” were reset. After the exposure is completed, the image processor receives a useable brightness-value from all photodiodes, including the ones which would have been saturated with a traditional sensor. Using the number of the “resets” and this final brightness-value, the processor can then completely restore the overexposed areas. The concept is called “Modulo Camera” and is based on a modular arithmetic. In the following video the way of functioning of the new sensors and the difference to conventional sensors are explained in an easily comprehensible way.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: