Google RAISR uses machine learning for smarter upsampling
 |
Top: Original, Bottom: RAISR super-resolved 2x, Image: Google |
Upsampling techniques to create larger versions of low-resolution images have been around for a long time – at least as long as TV detectives have been asking computers to 'enhance' images. Common linear methods fill in new pixels using simple and fixed combinations of nearby existing pixel values, but fail to increase image detail. The engineers at Google's research lab have now created a new way of upsampling images that achieves noticeably better results than the previously existing methods.
RAISR (Rapid and Accurate Image Super-Resolution) uses machine learning to train an algorithm using pairs of images, one low-resolution, the other with a high pixel count. RAISR creates filters that can recreate image detail that is comparable to the original, when applied to each pixel of a low-resolution image. Filters are trained according to edge features that are found in specific small areas of images, including edge direction, edge strength and how directional the egde is. The training process with a database of 10000 image pairs takes approximately an hour.
Once RAISR has been trained it is capable of selecting the most appropriate filter for each pixel in a given low-resolution image to fill in new pixels in order to create a higher-resolution version. RAISR can also remove aliasing artifacts, such as moiré patterns or jagged lines in the low-resolution images when creating the larger version, something that linear methods are not capable of doing.
Eventually, Google may be able to use the technology to upsample images that are sent at mobile bandwidth-friendly resolutions. More information on RAISR is available on the Google Research Blog.
 |
Left: Original, Right: RAISR super-resolved 3x, Image. Google |
|
|
|
|
Nov 13, 2019
|
Nov 6, 2019
|
Oct 21, 2019
|
Oct 16, 2019
|
The OM System M.Zuiko Digital ED 90mm F3.5 Macro IS PRO is finally here! Chris has opinions.
The Canon EOS R8 is the company's latest mid-level full-frame mirrorless camera. It brings the sensor and autofocus from the EOS R6 II and combines them in a smaller, more affordable body.
The Canon EOS R50 is an entry-level, company APS-C mirrorless camera. A 24MP RF-mount camera aiming to attract smartphone users and, perhaps, vloggers.
See what Chris and Jordan think of the Canon EOS R8 after shooting in Charleston, South Carolina.
The Canon EOS R50 is inexpensive. It's tiny. Is it good? Jordan and Chris dodged alligators in South Carolina to find out.
Above $2500 cameras tend to become increasingly specialized, making it difficult to select a 'best' option. We case our eye over the options costing more than $2500 but less than $4000, to find the best all-rounder.
There are a lot of photo/video cameras that have found a role as B-cameras on professional film productions or even A-cameras for amateur and independent productions. We've combed through the options and selected our two favorite cameras in this class.
What’s the best camera for around $2000? These capable cameras should be solid and well-built, have both the speed and focus to capture fast action and offer professional-level image quality. In this buying guide we’ve rounded up all the current interchangeable lens cameras costing around $2000 and recommended the best.
Family moments are precious and sometimes you want to capture that time spent with loved ones or friends in better quality than your phone can manage. We've selected a group of cameras that are easy to keep with you, and that can adapt to take photos wherever and whenever something memorable happens.
What's the best camera for shooting sports and action? Fast continuous shooting, reliable autofocus and great battery life are just three of the most important factors. In this buying guide we've rounded-up several great cameras for shooting sports and action, and recommended the best.
Comments