Filterless camera sensor tech from Panasonic may be a low-light breakthrough
The Achilles heel of contemporary camera sensors based found on the Bayer color filter array (CFA) is their reduction of light. Because every photosite is filtered to just obtain either red, green, or blue, over half of the light hitting the sensor is tossed away. For low-light instances, which represents the reduction of at minimum a stop of light — meaning cutting the camera’s shutter speed inside half.
Panasonic claims it has come up with a revolutionary hot method to redirect the light entering a sensor thus which most of it may be selected by the sensor. Instead of using an range of tiny microfilters inside a conventional CFA, the fresh approach utilizes what Panasonic calls “micro color splitters” which diffract the light thus which different combinations of wavelengths (colors) hit different photosites. In their paper inside Nature Photonics, Panasonics experts claim their answer enables the sensor to gather 1.85 occasions more light than conventional Bayer-array-based sensors.
Pioneers receive the arrows
Panasonic is not the initially business to attempt to throw off the yoke of the Bayer range. Foveon became well-known for its distinctive technology which utilized all light hitting its sensor by layering the 3 color receptors over 1 another. Each layer basically stripped off the color of light to that it was receptive, passing over the rest. Unfortunately for Foveon, they were alone inside utilizing this technique, thus all of the work of developing the sophisticated algorithms (plus hardware) for processing the resulting uniquely coded raw pictures was about them. As a outcome, it was years before Foveon had powerful sound reduction plus effective enough chips to make JPEGs inside the camera.
Similarly, photosites inside the Panasonic sensors won’t just register red, green, or blue. They’ll be getting combinations of colors: white+red, white-red, white+blue, plus white-blue, which come from the 2 deflectors used. Camera firmware should demosaic the results — calculating color values from those composites. Unlike with Bayer-array demosaicing — an active field of analysis with years of research behind it — Panasonic should pave the means for decoding pictures from its specific sensors. While it is actually potential to create RGB values with a easy matrix multiplication of the 4 color combinations received by diffraction-based sensors, doing it without sound is more complex.
Panasonic claims it has been capable to utilize the effective FDTD (finite differential time domain) way to compute the final image by simplifying it to be rapidly enough to compute inside real time. This offers a more exact image than simpler to calculate algorithms like B-BPM (Babinet beam propogation method). Like Foveon, Panasonic has protected its innovation with a thicket of patents. It has received or filed for 21 Japanese plus 16 overseas patents found on the technologies thus far.
Not truly the only attack found on the Bayer CFA
In a less-radical attack found on the Bayer range, Fujifilm’s high-end X-Pro 1 relies about a smart 6×6 arrangement of differently colored photosites to reduce the image artifacts found inside Bayer range cameras. Unlike Panasonic, though, Fujifilm’s approach doesn’t aid address the reduction of light matter. It nonetheless utilizes microfilters over every photosite that just permit 1 color to be recorded at every website.
Panasonic is reticent about a few of the key characteristics of its modern technologies. It doesn’t state what resolution is accomplished inside its click launch — though its experts have said which there is not a reduction of resolution compared to conventional solutions — or how much sound is introduced by the diffraction procedure. Those details can determine whether Panasonic’s approach is lucrative inside substituting Bayer arrays inside cameras plus smartphones or be relegated to specialized marketplace niches like surveillance which need the newest inside low-light performance.
Now read: The history of digital photography
[Image credit: Maxmax]


